The analysis of site audience
(using the data collected with the help of RCounter v1.0)
OR
17 useful conclusions, which
will help you to:
evaluate efficiency of your site promotion (as well
as your site general efficiency),
plan further promotion steps,
evaluate quality of your site structure,
point out its defects,
optimize your site structure,
and give you lots of other simply curious information…
First issue
September 18, 2001
Let’s begin this article with some questions, which we address to people who believe, that the number of site hits is only one thing, which is necessary to know about site audience:
- Having registered in the catalogs and search systems,
are you sure that the registration has passed successfully?
- If you used the program, which “registers a site
in 1000 catalogs worldwide” – are you sure, that if
only from such 100 catalogs somebody has gone to you?
- Having conducted a presentation, delivery or simply
having invited friends on your site, do you clear
understand effectiveness of this action?
- Are you sure the pages, which you diligently update,
are visited?
- Are you completely sure in correctness of your site
structure? Or the visitors do not go further than
the main page?
- Who is your visitor?
- In what days does your site have influx of visitors
and in what days does not? Are you sure, that in the
next time, when you will send your delivery, you will
not fall into “dead day”, when practically everyone
will ignore your letter, being busy with own problems?
- After all, do you strongly believe, that this list
of similar questions is completely definitive?
Contents
1. INTRODUCTION
2. BASIC PARAMETERS OF SITE ATTENDANCE
Conclusion 1. About how many pages of your site are viewed by the average visitor and as far as it’s good for you
Conclusion 2. About affection degree, constancy of your site audience
Conclusion 3. From which it becomes clear, how good your site “has sprouted up” on the Net
Conclusion 4. Describing “correctness” of your site main page
Conclusion 5. In which we analyze the actions of your site audience and demonstrate how to evaluate reasons of growth or falling in your site attendance
Conclusion 6. Assisting to understand, as far as justifies itself your site
3. ANALYSIS OF "REFERRALS"
Conclusion 7. About what part of your total site audience is the “devoted” one which probably being potential business partners if you have a corporate site
Conclusion 8. About “finding” your site in search systems
Conclusion 9. About what effect the registration in the catalogs has given to your site
4. ANALYSIS OF SEPARATE PAGES' ATTENDANCE
Conclusion 10. Devoted to site pages with abnormal low or high attendance
Conclusion 11. Assisting to evaluate quality of “key” pages
Conclusion 12. Sometimes permitting to detect idle (corrupt) links on your site
5. ANALYSIS OF THE CALENDAR STATISTICS
Conclusion 13. About efficiency of site promotion actions and on what days it is better refrain from such action
Conclusion 14. In which you will see the qualitative analysis of your site attendance surge as well as evaluation of conducted action
Conclusion 15. Again devoted to the dynamics of site audience changin
Conclusion 16. Which from time to time gives simply surprising results, essentially supplementing performances about audience interests of your site as well as about audience actions
Conclusion 17. Permitting to evaluate all way of visitors inside your site
Some other conclusions from the calendar statistics.
6. FINAL CONCLUSION
7. OTHER USEFUL DOCUMENTATION, INTERNET-RESOURCES, ADDRESSES
1.
Introduction
The analysis of site audience and furthermore – making
some conclusions on the basis of which is supposed to
improve site attendance - rather nontrivial problem.
Sometimes even a set of parameters of your site can
not give you clear picture of site attendance and sometimes
- on the contrary - everything becomes understandable
from “several digits”.
In this document I will share my own experience and
my colleagues’ one with you about how to make analysis
of site audience using RCounter product (www.rcounter.noonet.ru).
In the present moment this experience is based to the
analysis more than ten sites with the usage of RCounter
product (during 4-6 months), and about the same number
of sites until as we have developed this product.
I clear understand that many internet-analysts, to
whom probably you concern as well, have bigger experience,
however I pay your attention: the main purpose of the
document is to share the methods of researches with
reference to RCounter product with you (although this
document may be useful for users of other statistic
systems). Accordingly, you will find methods of rating
(analysis) of site audience mainly on the basis
of the data collected with the help of RCounter in this
document. Some parameters of site audience, which temporarily
are absent in the RCounter reports, will not be considered
in the given version of the document. But, undoubtedly,
they will be considered in the future editions, which
I will issue a little later, after appearance of the
new version of RCounter product (it will be later, because
me and my colleagues need enough time to work over and
check “new analysis technique”.
And the last one, what is
necessary to notice before the beginning of this informative
story – what the site audience was chosen as an example
to give you counsels and recommendations. I have selected
the web site of NooNet Internet-studio (www.noonet.ru)
for this purpose. Why have I choose exactly this site?
Not just because the audience of this site “is more
correct” than other sites have. And promotion/advertising
of this site in any way cannot be named model (exemplary-exponential).
Simply, this site from the moment of its creation has
been analyzed with the help of RCounter and its site
audience and structure are good enough for making understandable
examples on their base. Just look at this site as an
qualitative teaching and methodical manual, not ideal
but exponential one.
2.
Basic parameters of site attendance
In the beginning of this paragraph I’ll determine some
“elementary” parameters:
- Number of hits (showings of site pages);
- Number of hosts (visits);
- Number of unique visitors (IP-addresses);
- Number of showings of the site main page;
- Number of addresses-referrals (the length of “referrer
list”);
and some “more complex” parameters:
- Average number of “hits per day ”;
- Hosts per day;
- IP-addresses per day.
I have named them as elementary parameters, because
“RCounter” simply shows appropriate numbers on one of
the accounting pages and they do not need to be calculated
– just see them and everything will be understandable.
The version 1.0 does not show the more complex parameters
“obviously” and they need to be calculated on the basis
of other parameters.
The majority of the elementary parameters is visible
on the main page of RCounter:

Small preface about the basic parameters
It is not a secret that the more visitors the better
for this site owner. However, for sites of the different
type, size, regularity of updates (as well as different
quality) – different parameters will be good. For one
sites - 1000 hits per day is naturally, but for others
it is almost inaccessible level.
Therefore, we can not say, for example, 5000 hits per
day is good and 1000 hits per day is bad generally.
However we can make rather interesting and universal
conclusions on the basis of RCounter basic parameters.
Conclusion 1. About how many pages of your site are viewed by the average visitor and as far as it’s good for you
Let’s compare such parameters as “number of hits” and
“number of hosts”. It will be even better if we will
divide the first parameter into the second one. The
more result, the better for you. In my case (with NooNet’s
site) – the result is 1021/309 = 3.3. It means,
that the average visitor views more than three pages
of this site. Certainly, this result depends on subjects
of a site, number of pages. If your site has only 3
pages – it’s not possible to get this result, and if
your site has 100 pages, the given result is most likely
bad than good. However, in my case the size of my site
is 6 pages, therefore, this result is quite enough (because
the average visitor views more than half of site pages).
General rule: this parameter should be
not less than 2.2-2.5 for the majority of sites. For
some sites - even “3.3” will be bad result.
Be careful with conclusions: this parameter
is strongly depends on subjects and size of your site.
Conclusion 2. About affection degree, constancy of your site audience
We compare such parameters as “number of hosts” and
“number of IP-addresses”. Again it is not bad to divide
one into another. And, again – the more result, the
better for you. For my case it is 309/172 = 1.8.
It means, that on the average same person came on my
site twice for its existence history, besides - not
successively, but with the large interval (in different
days). In other words – he/she came back to my site.
The more percent of returns on your site, the better,
because it shows “affection degree” of audience to your
site. And such parameter is always lowest on resources,
to which the visitors get “casually” (via search engine
or banner exchange system).
It is necessary to make reservation that “number of
IP-addresses” inaccuracy shows real number of different
people coming on your site. Under an one address the
different people (for example from one company with
proxy) could come, and on the contrary - the same person
could come under different addresses (from different
computers or with dial-up connection). But, on the whole,
such parameter as “number of IP-addresses” can be used
in the audience analysis. It will not be a big mistake.
General rule: the given parameter is
not bad if it is more, than 1.4 and it is very good
at values more, than 1.8. If for your site this parameter
is “2” and more - it means, that your site has “especially
devoted” audience, that meets rather seldom.
Be careful with conclusions: do not try
to analyze this parameter when your site is too young
(less than 3 months). At the early stage of site existence
it is impossible to say, that its audience is devoted
or chance – your site audience has not formed yet.
Not recommended conclusion: After reading conclusions
¹1 and ¹2 you may desire to receive the relation of
number of hits to number of IP-addresses. In other words,
to find out how many shows one unique visitor has seen
on average. I would not recommend considering such parameter
and making any conclusions with its help. It is not
only difficult, but also is very inaccuracy. You see
it is not known from this parameter, when the same visitor
“stroll” about your site, and when, with big breaks,
came back to the same page. Therefore it is easier to
make an incorrect conclusion, than receive any really
curious appraisal.
If you disagree with me and see good practical usage
for this parameter – I’ll familiarize with your interpretation
of this parameter with great pleasure (send your letters
to amfora@lvs.ru).
Conclusion 3. From which it becomes clear, how good your site “has sprouted up” on the Net
Let's look at number of addresses-referrals (URL count
in Referrer list). This parameter does not need to be
divided into anything, simply take away 2 addresses
(because two addresses are always your site and “unknown”
in this list) from it and estimate obtained result.
For me - 19-2 = 17. As a rule, this parameter
depends on intensity and breadth of advertising/promotion
of your site as well as its lifetime
General rule: Less than 50 referrals
for a site living 3 months or longer and past registration
in the catalogs/search engines is not enough. By the
way, my example - NooNet is a Russian web-site, so it
is not lawful to use these criterions for this site.
More than 100 referrals is good even for a site “in
its heyday” (for one year or older sites). And if you
have more than 200 referrals – it is simply excellent
parameter. May be you have achieved it not in the usual
manner.
Be careful with conclusions: Pay attention
that number of referrals has different value for sites
of different age, with different intensity of promotion
etc. But greater importance is what geographical specialization
of your site - international or regional.
Conclusion 4. Describing “correctness” of your site main page
Firstly, count how many percents from total shows (hits)
the main page shows makes. In my case (352 / 1021) 100
= 34 %. This number obviously shows what percent
of your visitors past further than the main page. Than
less this number, the better. When this number is great
enough - always it is necessary to think about it: may
be something wrong in your site structure or in presentation/registration
of the main page?
On the first sight, this parameter is similar to percent
of hosts from total hits (If we suppose that a visitor
browses through your site - that necessarily one hit
belongs to the main page, and others – to the second
pages of your site). But such assumption is very inaccurately,
because often visitors can load (update) the main page
some times or periodically to come back during their
browsing through your site. And, on the contrary, a
visitor can get on your site and browses through it
not enter into the main page. Thus, the percent of the
main page shows from total hits is an independent valuable
parameter.
General rule: If “the percent of main
page” is more than 45 % - it almost always shows that
a site has problems. This “percent” should be 30-35
% and less for the majority of sites. If this value
is less than 25 % - it is already very good.
Be careful with conclusions: the value
of this parameter strongly depends on both the subjects/size
of your site and the type of your main page. There are
main pages, which have a lot of information (on which
a visitor frequently receives everything what he/she
wants and goes away) and few-informative ones (which
are only “for meeting” a visitor). In the first case
“the percent of main page” may be big enough, in the
second case – it must be rather small.
Conclusion 5. In which we analyze the actions of your site audience and demonstrate how to evaluate reasons of growth or falling in your site attendance
Firstly, calculate the average number of “hits per
day” and “hosts per day”. Compare these parameters with
the current daytime attendance of your site (using parameters
of one day only is very inaccurately, therefore, you
can take the average level of attendance under the “totals”
of current or previous month for “current attendance”,
see the calendar statistics).
The conclusions by results of such comparison
can be made more interesting than it seems on the first
sight.
If the “current” number of hits/hosts per day
is more than “average for site existence”, besides they
have grown approximately proportionally - then the conclusion
is simple - the audience of your site grows. If these
are proportionally less – your audience reduced. But
if there is disproportion (for example, the level of
hits grows but the level of hosts does not) it is necessary
to make more complex analyses with this information.
When the growth of hits overtakes the growth of hosts
– we can say that your site audience approximately stays
on the same level, but your visitors go to your site
with bigger enthusiasm; in the case when the growth
of hits are behind – your audience grows, but most likely
the real reason of this is some special methods of promotion
and the majority of new visitors has low interest to
your site. If you see some drop of hits/hosts – it means
that the disproportion also has some special feature:
when hits fall faster – it means that interest to your
site is lost (for example, all information has already
read, so visitors simply check your site for some news
or upgrades); when hosts fall faster than hits do –
it frequently means that one of strong promotion ways
which gave you lots of new visitors is getting ineffective
but old and devoted part of your audience remains and
it creates comparatively high level of hits.
The most ambiguous parameter is growth of hosts
at falling hits or there is reverse situation. It’s
most likely temporary status meaning intensive changes
in your audience, which definitely should be replaced
with general growth (both hits and hosts) or (alas!)
general falling.
General rule:
Practically we cannot say that for example “so-and-so
level of hits/hosts growth” is normal but “so-and-so
one” is too small. It depends on how active you advertise
(promote) your site. For a site with no updating/promotion
even “zero balance” is a good parameter. On the contrary,
for an actively promoted site - the monthly audience
growth in 10 % most likely is a negative result. However,
from my experience I can say that the monthly growth
of hits/hosts should be 2-3% for a typical “live” site
with stable audience anyway. It happens due to: a) general
Internet growth; b) dying of competing sites. “Zero
balance” speaks most likely about neglect than stability
of your site. Though, it’s only my observations of some
cases and I do not want to say that it works definitely
in all cases and situations.
Be careful with conclusions:
When you evaluate the level of hits/hosts growth or
fall always remember that this parameter has a significance
only against a background of that advertising (marketing)
campaign, which you conduct on the Net as well as against
a background of campaigns of competing sites. Because
it may be that everything is O’K with your site but
your competitor began very actively “to gather” additional
audience, well … in any sense this situation can be
considered as a negative parameter, which shows your
gap from this rival. And don’t forget that there are
seasonal fluctuations in the Net attendance as a whole
and your site personally, for example, corporate sites,
as a rule, are visited a little in summer as well as
entertaining ones during student session.
Be careful with young and small! Do not
try seriously to make ratings of growth or fall for
hit/host level for young sites, which did not have constant
audience yet. If your site lives less than 3 months
– all conclusions most likely will be very doubtful
- less 6 month age is also questionable. This method
of analysis works well for sites living more half a
year or greater. But it is not the last limitation!
Also, pay attention when you use the described method
for sites with low level of attendance. Just remember
that 10, 15, 20 are not statistical parameters. For
qualitative statistics - it is good to have not less
than 100 hits and 30 hosts per day then the analysis
will have real sense. By the way, because of low level
of attendance I did not begin to describe this method
using NooNet site. If I exampled on my “5-6 hosts and
10-20 hits” - any skilled analyst completely fairly
would say that I am not right, gently speaking.
Conclusion 6. Assisting to understand, as far as justifies itself your site
Firstly, calculate the average number of unique IP-addresses
per day. In my case it is 175/101 = 1.7. This
parameter should be considered as: daily 1.7 new visitors
on average find out your site – someone simply learns
and goes away, others see your site in detail. If you
have a corporate site – just image – that new potential
clients visit your office. If you have a site of other
type - conduct any suitable analogies. This parameter
should force you to think about your site profitability
and necessary to strengthen its advertising/promotion.
General rule: For a good corporate site,
which has not intensively promotion on the Net, about
20 new unique visitors per day can be considered normal.
But it also depends on site subject (company specialization).
For informational, entertaining sites it is difficult
to give universal ratings (for one it’s good to have
50, for others even 200 is not enough).
Be careful with conclusions: First of
all, look at this parameter extremely against a background
of expenses on your site development/tech.support/promotion.
For a site, which “lives without big expenses” sometimes
10 ip/days can be considered as an acceptable level.
But for a serious Internet-project, which constantly
needs separate employee and some finance for promotion,
100 ip/days cannot be considered as a satisfactory parameter.
3.
Analysis of “referrals”"
I think it is necessary to devote the separate chapter
to analysis of the "where from" statistics or - “referrals”.
“RCounter” shows this kind of statistics on the separate
report page. It looks in the following way:

One conclusion, which can be received, looking at the
number of addresses in this list, was already described
above (see conclusion ¹3). However, we can receive more
interesting conclusions from the "where from" statistics.
But, firstly, I should say about:
Problem of frames in the statistics of referrals. If your site is built
with the help of the frame technology, the statistics
of referrals will be very inaccurate and may be incorrect
at all.
It occurs for the following reasons - when a visitor
comes on a usual site (without any frames), at the first
transition the address of the site from where he/she
came is taken into account as a referral. When the visitor
“wanders” inside the site – the site address is taken
into account as a referral. Thus, all addresses of referrals,
distinct from site addresses, are addresses of other
resources from which the transition is carried out to
your site. But when a site is built with the help of
frames every time “primary source site” address is taken
into account as a referral at the first transition and
during "wander" inside a site quality referral is set
off. But it is not correct.
I'll explain it using an example: If I have found a
site "X" with the help of “AltaVista” search system,
passed on it and thumbed through 4 pages of this site,
then if the site is built without frames – 1 transition
with “AltaVista” and 3 transitions with “X” will be
registered (total – 4). It is correct. If the site is
built with frames - then all 4 transitions with “AltaVista”
will be registered - that is completely incorrect.
“Frames” is another reason why I have taken “NooNet”
site as an example for our analysis. It is built without
frames and consequently its statistics of referrals
is good enough and correct. Let's see it.
Conclusion 7. About
what part of your total site audience is the “devoted”
one which probably being potential business partners
if you have a corporate site
First of all count, how many percents from all transitions
the transitions in the “unknown” line make. In my case
it is (276/381)100 = 72 % (381 is the sum of
all lines except for the first, because the transitions
with “NOONET.RU” are transitions inside the site. They
are not interesting to us). The higher obtained percent
- the better for you. You see that the “unknown” transitions
generally are references to the site either with the
help of the link from a letter/document or directly
inputting the site name in the address bar of a browser.
But, in any case - the visitors who have come on your
site “from nowhere” usually are very loyal members of
your audience, who come on your site not by chance and
probably it is not for the first time. So, the main
thing for corporate sites is that high percent of “unknown”
means high percent of business partners (even potential),
whom your web-site attracts.
I say honestly that “unknown-visitors” are sometimes
registered and by virtue of technical reasons (when
it is impossible to determine the referral address)
but it is not so often case and we can consider this
factor as slight error of gathered statistics.
General rule: It is good to have this
parameter greater than 50% for any site. However, really
good result is 70-75 % and above. If this parameter
is 40 % and less - your site is visited mainly thanks
to good links from catalogs, good place in search systems,
banner exchange systems and so on.
Be careful with conclusions: Above I
said that “unknown” sometimes can be take into account
from “technical reasons”. Again I say honestly – the
nature of these reasons is comparatively difficult and
I am not sure that it absolutely always is a “non-significant”
factor. Be careful - if the “unknown” level is improbably
high on your site – it is good to clear up reasons of
it, for example, by practical way (independently coming
on your site from “anywhere” and from search systems
look at correctness of your statistics), or with the
help of scientific methods (ask experts about it).
Conclusion 8. About “finding” your site in search systems
At first, count how many percents from all transitions
the transitions in lines with the names of leading search
systems (AltaVista, Google, HotBot) make. In my case
the considerable results are gotten only from AltaVista
system: (22/381) 100 = 5.8 %. If your site has considerable
percent of transitions from either one or another search
system – it means that your site is “well prepared”
for that search system robots. If this percent is insignificant
or there are no transitions at all – your site is badly
prepared for the given search system.
Please, pay attention that this conclusion shows only
the level of “finding” for your site in search systems.
If you are interested in good efficiency from Google/AltaVista/HotBot
or other search systems – this parameter is of great
value for analyses of your statistics as well as it
will allow you to optimize your site for search robots
better. But not always site “finding” in search systems
is really valuable characteristic. In my case this described
parameter will have only informative meaning.
By the way, the case when the site “is searched” well
in one search system (for example, AltaVista) and bad
in other one is not rare and not surprising. The fact
is that the various search robots process search and
ranking of sites, being based on different criteria
(one - on contents of page, others - on keywords, header
or use a combination from those and other parameters
of a site). Therefore good “finding” in AltaVista and
the absence of such result in Google show that part
of a site, on which Google is oriented, is not worked.
General rule: For sites of different
subjects “normal percent of search system” also will
be various. In any case, less than 1 % of transitions
is an unsatisfactory result. More than 6-8 % is good
anyway (Here I mean a percent of each search system,
not their sum!). However there are cases when more than
15 % of transitions on your site is given by any search
system, moreover, there are sites, which in general
“gather” their audience mainly with the help of search
systems. For such sites, certainly, this percent should
be 15-25 %.
It is necessary also to notice that, as a rule, AltaVista
gives more visitors than some other system – only therefore
that AltaVista is more popular and often used system.
So, do not wait that a percent of transitions from all
systems will be equal. However, sometimes it happens
that by virtue of specificity (or subject) the site
is visited from Google or HotBot more often.
Be careful with conclusions: For young
sites the absence of any result from either one or another
search system can mean that the robot of this system
simply had not time to check up and index your site.
As a rule, the period of search and indexation of new
sites for search systems is about one month.
Conclusion 9. About what effect the registration in the catalogs has given to your site
Firstly, determine for yourself the list of Internet-catalogs
(key sites etc.), which you wish to see as large sources
of your visitors. Probably, you sent the applications
for site registration to administrators of all these
resources. Now, check up from what resources the visitors
came and from what ones do not. Count the percent of
calls for resources from which the visitors have already
came as it was made with search systems.
The conclusions concerning successes of registration
in the catalogs (key resources) are made similarly to
“finding” rating in search systems. But the conclusions
will be a little bit different. If for a search system
is mainly important qualitative construction of keywords
and text of page header, for the catalogs it is important
qualitative description of your resource and placing
the link on it in a “correct” rubric.
If you see very small percent of calls from any catalogs
- it is possible to assume that the link to your site
stands not in “correct” rubric, on an improper place
or the description of your site has appeared unsuccessful
and it does not convince a visitor to go on your page.
If you see no visits from a catalog, probably, the link
to your site was not added at all. The reason can be
an incorrect filling of catalog registration form or
moderator’s decision (a manager of the catalog) about
lack of correspondence of your site. Anyway, you should
think about the correctness of your site registration
again, most likely, you have to repeat the procedure
of registration in the catalog (make a re-registration).
Pay attention! Be careful when you consider that this
catalog gives you “enough” or “few” visitors. Do not
forget that various catalogs themselves have very different
attendance. So, it not surprise when, for instance,
such large catalog as “Yahoo” can give you more visitors
than “AltaVista”, but any modest
narrow-thematic catalog simply can not do it. Nevertheless
the visitors coming from that narrow-thematic catalog,
probably, are more valuable for you than visitors from
“AltaVista” or “Yahoo”.
By the way, it is very important for attraction more
visitors with the help of a catalog to have a good place
of the link to your site and its decoration. Some catalogs
have paid services for placing links on an attractive
place (for example, in the beginning of the list, outside
of an alphabetic sequence) or some service in highlight
of the links by the font/color/frame. Probably, it would
be better for you to "experiment” with such service
for a short time to evaluate how many additional visitors
from this catalog you got using this service. After
that you can decide for yourselves whether it is necessary
to spend finance such “favorable presence in this catalog”
or it does not justify itself (may be you can spend
this financial resources on other kinds of promotion
on the Net better).
General rule: relatively large Internet-catalogs
(Yahoo and some others) – in some cases can give you
more visitors than search systems - that is the percent
of transitions from them can make 5-10 %, sometimes
– even more. For specialized resources (including branch,
corporate) large percent can be given by an advanced
branch, regional catalogs (5-10 % and more) it occurs
because, that it is often users look for specialized
resources not in search systems but in thematic catalogs
(for example, a resource about programming is very convenient
to look for in the appropriate section of “Yahoo”, where
necessary subsections are created and you can find qualitative
descriptions instead of rummaging in the mammoth list
of found documents, which "AltaVista" gave on the word
“programming”).
Be careful with conclusions: Once again
it is necessary to remind that low level of your visitors
from either one or another catalog can be not only unsuccessful
registration (unsuccessful description) but also simply
low attendance of a catalog or (more often) - low attendance
of a concrete rubric of either one or another catalog.
And do not forget that registration in a catalog is
not necessarily made in the same day then you fill in
the application. All the more, it is impossible to wait
for new visitors from this catalog in the same day.
However, catalogs give faster effect than search systems
– you will get new visitors within 1-2 weeks after successful
registration. If a catalog makes deliveries of news
about the new added resources – in this case, you can
have a big surge of attendance even within several days
after registration.
4. Analysis of separate pages’ attendance
It has been already mentioned about the analysis of
the main page's attendance before (conclusion ¹4). But
the attendance level of other pages also can be very
important for understanding how good your site (its
structure and contents) satisfies visitors' expectations,
what exactly is interesting for them on your site, and
what pages are abandoned by visitors.
The RCounter program shows the attendance statistics
of separate pages in the following way:

Now we are ready to make conclusions, which can be
made, looking on this statistics.
Conclusion 10. Devoted
to site pages with abnormal low or high attendance
When this list is sorted out by number of shows of
pages, as in my example (the sorting mode is adjusted
in the program) - the order how pages follow in the
list corresponds to the level of their readership -
though to be precise – perhaps not the level of readership
but number of shows for all page history. Thus a very
popular but young page can appear in the bottom of the
list and poorly readable but old can appear above. However
you, as an owner or editor of your site, certainly,
know everything and, looking at the list, can easily
evaluate whether there are no “anomalies” in it.
To my mind it is the most useful conclusion, which
can be made looking at this statistics. I'll explain
how I understand “anomalies” slightly more in detail.
When a site page, which has a very modest role and which,
probably formed without a special technology, appears
too highly in this list with a big separation from others,
more “important” pages - it is an anomaly. On the contrary,
when a very important page, for which you (or your colleagues)
spent a lot of time for its construction, design, checking
and you hoped that its contents will reach many visitors,
but it appears in the bottom of the list (or even in
the middle but with the large backlog from “leaders”)
- it is also an anomaly. There is another anomaly -
for instance, when a press release, an announcement
or simply a new publication, which you widely “advertise”,
starts slowly instead of “big jump upward”. It begins,
like a “simple” page, to creep in the list.
In my case, there is only one anomaly (I see it as
an editor of this site) – it is very low “rating” of
the page “Horizont”, where my article is located. I
have spent a lot of time for article writing and expected
that it will be rather interesting and useful to this
site audience. It does not happen - and this fact gives
me a cause to think about it.
What does the presence of anomalies for a creator (editor),
owner of a site mean? The reasons of anomalies can be
various. Below, I located the main ones:
- A link to an “abnormal unpopular” page, its description
(comment) is not clear for your visitors. It does
not give correct information about its contents. I
believe that this is the source of my anomaly.
- A link to an "abnormal unpopular" page is located
in an unsuccessful place or gets lost among other
links (for example, at the image resolution 800x600
– it falls outside the screen etc.).
- You know badly interests and passions of your audience.
For example, if students visit your site - they actively
will look for free resources and articles and such
pages, as price-lists will be interested seldom by
them. If your visitors - businessmen, they will be
interested in business offers, prices, analytical
articles, and only afterwards scientific publications.
- An “abnormal readable page” is popular due to good,
attractive description, successful keywords (which
people often use in search systems). May be its description
(keywords) is not correspond to contents. So, visitors
go with the expectations of receiving something on
this page but they definitely see another thing -
and it is too poorly (despite of its high attendance).
- Also a page may be “abnormal popular” because of
special placing of its link. Do not forget that when
a visitor has some difficulties to select interested
page he/she does not leave from your site immediately
– a visitor presses most likely the first link in
the list or the most observable and so on.
- When a new and “promoted” page (a press release
or a special business offer) very languidly attracts
visits – the reason, probably, is its misplace of
the link to this page or simply inertness of potential
visitors. For instance, if you have delivered a letter
"please, get acquainted with the special offer, which
is valid to the end of this month" to all partners
– it stands to reasons that not everyone will look
for this offer on your site. It would be a great idea
if you put the link to this page directly in the letter
as well as highlight the appropriate link on your
site (font, color, frame and so).
- Also “slackness” of a new and promoted page may
be from your inexact insight of interests of your
site audience. May be they do not interest “news”
or “offer” at all.
But this list of possible reasons of “anomalies”, certainly,
is partial as well as not so concrete. At detection
of “anomalies” your experience, common sense, and, probably,
consultation of the experts will be play the key role.
I would recommend to ask some people, who are not involved
in your project at all, why this page attracts redundant
interest or, on the contrary, is not popular to their
mind. Often such kind of people can easily find defects
of your site, which you don't see or miss accidentally.
General rule: Generally speaking, the
above list was composed on the basis of my experience,
but, as usual, I'll say some words in this rubric as
well. Most often, owners and editors of sites are worrying
about abnormal low parameters either one or another
pages. In the overwhelming majority of cases the reason
is bad structure of a site, unsuccessfully formulated
links, unsuccessful location of links. Much less often
(from my experience), but also you can meet - a mismatch
of a page contents with needs of visitors, in other
words – page creators simply inaccurate know the audience
interests.
Be careful with conclusions: It is necessary
again to remind you that total of page shows is displayed
in this kind of statistics. Naturally, old pages will
be above young ones, even if they are more popular.
The current attendance of pages you can see in the section
“calendar statistics”. In the future RCounter version
it is planned, specially, to show their average popularity
in the report about attendance of pages.
Conclusion 11. Assisting to evaluate quality of “key” pages
Now, special attention should be paid on attendance
of the separate type of pages - “key” pages, from which
there are links to other pages. The special attention
should be given them for the reason that the popularity
of pages to which they refer strongly depends on them.
Unfortunately, in my example (NooNet) there are no such
pages except the main. But for many sites such pages
are pages with the list of articles/publications or
they are production catalogs from which it is possible
to get on pages of separate goods or something similar.
What special conclusions can be made, considering their
attendance?
Firstly, the attendance of key pages should be high
enough. As a rule, a site should have the best attendance
on the main page, further - on all “key” ones, and only
after that on separate informative pages. Though, there
are exceptions in this rule. But if the attendance of
a “key” page is lower than for some “simple” one – it
is a good reason to think about it. You see, a “key”
page is, conditionally speaking, a “crossroad” from
which there is a set of roads and which should differ
by rather high popularity. Moreover, sometimes, it is
a good idea to register especially “key” pages in catalogs
and make special decoration of it for search systems
– they can have mammoth independent value.
Secondly, you should compare the “key” page attendance
with the attendance of pages on which it leads. If the
attendance of “wards” pages in the sum is too exceeds
(thrice as large and more)
the "key" page attendance that it almost means that
“final” pages get visitors basically with the help of
search systems, instead of the “key” page of your site.
Sometimes it works. But, sometimes, it testifies that
your “key” page is visited too little or there are unsuccessfully
indicated and commented links to other pages on it –
in this case, visitors simply leave your site without
trying to clear up with your site contents.
By the way, the main page of your site can also be
considered as a “key” page and you can apply all above
judgements as well.
General rule: I should say that the conclusions,
described above, are not invention at all and their
advantage is quite real. I repeatedly meet cases, when
“key” pages and even the main pages of sites badly distribute
visitors on “sponsored” pages. Moreover, there are “key”
pages, which even now do not work on the high-powered
level on some sites, where I am an editor – but I am
going to spend a lot of time, thinking about the reasons
of it, experimenting, and optimizing structure of these
sites. This, I advise you too.
Be careful with conclusions: do not forget,
sometimes, it happens that some “final” page is visited
more than the “key” page or even the main one. It happens
when there is a very popular resource (an article, downloading
file) on such page and there are a lot of users from
catalogs and search systems who get directly to it.
It is an exclusive case and it should be considered
as exclusive one, do not try that the “key” page attendance,
certainly, overlaps the attendance of this “star” page.
There are also other cases, when the discourse logic,
described in this conclusion, should be applied with
the certain clauses. Follow by your common sense, analyze
statistics, and make experiments!
Conclusion 12. Sometimes permitting to detect idle (corrupt) links on your site
This conclusion is rather simple, but sometimes extremely
useful. Look at the date of last visit of pages. Even
better – choose the mode which sorts the list of pages
by date/time of their last visit from the RCounter program,
then you'll see pages with the longest time without
visits in the bottom of the list. Now, ask yourself:
"Are there such pages, which nobody visited too long
ago?". Such a long absence of visitors on a page
can mean that its link “has broken”. May be you or your
colleague casually have broken, editing the site, or
deleted the link to this page. May be something else.
When you see such kind of pages - simply go on the
site and check up their accessibility.
General rule: You may think that this
conclusion practically never is useful. However for
relatively short time of RCounter usage I have already
found out twice such “dead” pages. Once – a link was
casually corrupted by web-master, in other time - I
noticed at once some “unvisited” pages and remembered
that their links never placed on the site. Because,
a long time ago, I was going to create a separate rubric
for this purpose, but for my shame, I forgot.
Be careful with conclusions: Generally
you can be careless with this conclusion. No problems,
if you once again suspect “link breakage” and at checking
it will appear that everything is OK - the page has
simply low popularity. It would be another reason to
think again - why popularity of a page is so low.
5.
Analysis of the calendar statistics
The calendar statistics is the information about distribution
of visits of all site and pages by days. I can say that
it is almost the most informative and useful part of
the RCounter reports. Probably, it should be considered
in earlier chapters, but if you're reading the document
here – it means that you are really seriously interested
in the attendance analysis of your site. Well, forthcoming
conclusions from the analysis of the calendar reports,
about which I'll tell you, now, will be an excellent
dessert to all aforesaid…
As usual, I give you an example of the RCounter statistical
report:
again I return to conclusions:
Conclusion 13. About
efficiency of site promotion actions and on what days
it is better refrain from such action
Look at the lines “Total” and “Hosts”, selected in
the report by a light-blue background. There is the
information about all site attendance, by calendar days
in them. First of all, we interest in surges and collapses
in attendance. For my example – there are one strong
surge (9th, 16 hosts/48 hits) and one obvious
collapse (7th, no visits at all).
Surges practically are always concerned with some site
promotion actions on the Net or promotion actions on
the offline (presentations, exhibitions, conferences)
and also - that someone, independently from you, could
publish the information about your site (in a journal
or Internet-review). Furthermore, for example, an Internet-catalog
could independently find your resource and add it in
the rubric “new Net resources” or mention about your
site in deliveries.
When you note any surges, it is always necessary to
ask yourself – what is the reason? And, on the contrary,
making some advertising actions (on the Network or offline)
- it is always necessary to look at whether the attendance
surge is after several days and how big it is. Finding
out natural surges (after your actions) it is necessary
to evaluate how well such actions hereinafter; observing
surges which are independent from you - it is necessary
to understand their reason (if it is possible) and think
about how you can initiate them in the future.
The separate subject is collapses in attendance. As
a rule, they are called by week or seasonal attendance
fluctuations, concerned with “outside the Net” activity
of your visitors. For corporate sites - collapses are
natural on weekends, for sites with abstracts - during
student vacations and so on. But, it is necessary not
only to predict collapses, followed by common sense,
but also to see what occurs with attendance in the reality.
If you see that there is an attendance collapse each
Monday on your site, it probably means that the majority
of your visitors comes at office and have a lot of businesses
"outside the Net" on Monday. They simply do not have
time and need to go to your site on Monday. The same
is true about the analysis of “surges”.
You can make especially practical conclusions (namely,
about possibility and expediency of repeated initiation
of such surges). The similar practical conclusions are
possible for the analysis of collapses as well. More
concretely, the analysis of collapses allows you to
plan the promotion actions of your resource (advertising,
presentations, deliveries) better. You can avoid making
such actions during the periods of natural-low activity
of your audience, when those visitors, who you expect,
will poorly remark these actions.
General rule: the overwhelming majority
of attendance surges can be detected easily. As a rule,
they are concerned with promotion actions of your site
collective or offline actions. Even simple listing and
classification of these reasons/actions deserve separate
articles and books devoted to promotion of resources
on the Net. As to collapses
in site attendance, as a rule, it is easy to select
their reasons and almost always it is concerned with
week or seasonal periods of visitor activity.
It is difficult to overestimate the real practical
value of observation for attendance surges and collapses.
These observations have a great value for correct planning,
which allow you to intensify effect from these actions
and, simultaneously, reduce the costs – simply refuse
from small-efficient actions or reduce their size.
Be careful with conclusions: Analyzing
surges and collapses in site attendance, sometimes,
it is possible to make incorrect conclusions. Comparatively
surges – it is can happen when some actions (offline
presentation and delivery on the Net) are simultaneously
used. And it is difficult to understand what action
gives observable effect. Comparatively collapses – do
not forget that they can be spontaneous, especially
if your site attendance has a low level (less than 100
hits per day). Both in that case and the other one the
insurance from errors should be the common sense, recommendations
and advises of experts, and regular practical confirmation
of your observations as well.
Conclusion 14. In which you will see the qualitative analysis of your site attendance surge as well as evaluation of conducted action
Let's finish the subject of analysis of attendance
surges. The separate conclusion can be made not in the
area of searching surges and analysis of their nature,
but researching their characteristics. A surge is characterized:
by number of hits/hosts, duration and schedule of fading
(how intensive your site is visited in following days)
and what pages have visit surge. The analysis of these
parameters allows you to understand what visitor has
gone on your site – the same, as earlier, or a little
other one. This conclusion is the additional help in
the evaluation of conducted action and in determination
what effect you achieved and what you can get from this
effect (increasing devoted audience or even new commercial
partners; but they may be simply temporary visitors).
Now, I’ll give you an example. In my case - in the
peak it is registered 48 hits and 16 hosts. In other
words, the relation of hits/hosts is 3. It is approximately
equal to similar parameter for all site pages. So, I
can say that the degree of the interested audience,
which has come on my site during this surge, is similar
to common parameters. If hits were less – it means that
less interested audience has come than usual site audience
during the surge. If hits were more - more interested
audience (so, the promotion action can be considered
useful not only in a quantitative increase of audience,
but also in qualitative one).
As to duration of surge - for me, it has a strong expressed
view for one day, though, there were 3 days when attendance
was hardly above the average level after this day. Here,
it is impossible to make unambiguous conclusions, but
there is an impression that it is also concerned with
surge. So, the conducted action had an effect on the
site attendance about 4 days. It is longer than the
effect from usual “Spam deliveries”. It is approximately
equal to the effect duration from one offline-presentation,
but, obviously, it does not correspond to the effect
received from some day exhibition. Besides, this surge
does not correspond to the effect which can be achieved
at registration in Internet-catalogs, as the surge is
either more shortly (1-2 days, when registration passes
in small-sized catalogs) or it is more mark (not 16
hosts, but at least 30, when there is a registration
in large catalogs).
Making the analysis on what pages visitors came during
surge, I find that as a whole “distribution by pages”
is standard for my site audience, though, on the pages
“Horizont” and “Addresses” the percent of visits was
hardly above usual. Most likely the reason is that the
site has been gotten new visitors. And many of these
visitors browsed different pages for better understanding
of the site.
General rule: Repeatedly having observed
results of either one or other advertising measures
and promotion of sites I strongly believe that considering
the characteristics of surge, it is possible to understand
a lots relatively the quality of new visitors, and consequently
– relatively the efficiency of the conducted action
both in quantitative and in qualitative aspect. Often
you can see a peak for surges per the first day and
then there is more or less sharp recession to the average
level attendance. Sometimes it happens differently.
Generally speaking, discourses about for what types
of measures, what kind of attendance surges are typical
– are a separate subject and, now, my experience in
this area is not so great to enter any classification,
find regularities and give "big" advises. Probably,
I'll be able to supplement considerably this section
or even create a separate document about this subject
with the help of your remarks and comments.
Be careful with conclusions: the value
of surge is an important parameter for the analysis
of action effectiveness. As I also work with statistics
here, so values should be at least statistical. Making
conclusions relatively my 16 hosts, 48 hits, and 9 visitors
of any page, I am not quite right. But, I tried to show
a principle. Furthermore, when number of hits/hosts
in the peak exceeds thrice as large the average site
attendance level - it is already sufficient parameter
for some conclusions. But, of cause, we should try to
make serious conclusions on the basis of serious parameters
- hundred hits/hosts, tens shows of different pages.
Do not forget that you can try to make either one or
other conclusions always, but not always they can be
made with an adequate accuracy and have a high degree
of reliability.
Conclusion 15. Again devoted to the dynamics of site audience changing
It is possible to receive good enough parameters of
site attendance growth or falling from the calendar
statistics. For this purpose, compare values in the
cells Total/Average and Hosts/Average
(the average number of hits and hosts per day, accordingly)
with the appropriate indications for the last month.
A one remark – this comparing can be made not earlier
the fist decade of a month, because the statistics for
some days may be very incorrect.
If the average number of hits/hosts has grown in comparison
with the previous month – it means that the attendance
grows and on the contrary. In my case – 12.39 hits/day
and 4.17 hosts/day are fixed for the current month.
For the previous month - 9.07 hits and 3.1 hosts per
day. You can see that the attendance has grown well
(approximately 1.36 times as many). Thus, the percentage
ratio of hits/hosts remained approximately the same
as for the last month. It means that the qualitative
changes in the audience have not taken place only the
simple quantitative have taken place.
See the conclusion ¹5 for detail information observing
growth/falling of hits/hosts. But there is the comparison
of the current number of hits/hosts with the average
parameters for all site pages existence in the conclusion
¹5. It is possible to make a comparison of current month
with previous or any other one (for instance, with the
same month of the last year) with the help of the calendar
statistics. Thus, it is possible to make considerably
more complete performance about the site audience actions.
General rule: Just see the appropriate
section of the conclusion ¹ 5.
Be careful with conclusions: It is necessary
to warn against hasty conclusions about growth or reduction
of site audience under the observation of the calendar
statistics. If there is a comparison of current parameters
with the average ones for all site existence in the
conclusion ¹ 5, here – I use month parameters only.
It can give incorrect result, when the last month had
a period of natural recession of activity, and the current
one has a period of natural growth (for example, the
last month - August, and the current one - September).
And, on the contrary, when the last month had a period
of growth, and the current one has a period of recession
– using this conclusion is not quite right, as the fall
of hits/hosts level does not show that the site audience
is reduced in this case.
I recommend using the discourses described in this
conclusion, together with discourses of the conclusion
¹ 5. Together they will give clearer performance about
the actions of your site audience.
Conclusion 16. Which from time to time gives simply surprising results, essentially supplementing performances about audience interests of your site as well as about audience actions
After you have evaluated growth or falling attendance
of your site – it is time to use unique possibilities
of the calendar statistics - to find out what pages
people visit or on what pages visitors have stopped
to visit.
For this purpose there is the column “Average”, indicating
the average number of hits per day for each page. The
same column is in the report for the last month. It
allows you to compare these parameters for each page,
so you can definitely give the answer to the delivered
question.
In my case, the following parameters of pages are fixed:
Addresses = 0.89, Development = 2.33, Horizont = 0.61,
Info_n_News = 2.22, Main = 5.11, Participants = 1.22
for current month.
For the last month the similar parameters have values:
0.70, 2.03, 0.40, 1.07, 3.40, 1.43 (accordingly).
Now, it is necessary to compare these numbers, paying
attention for the general parameter of site attendance
growth/falling. I remind you that the general site attendance
in comparison with the previous month has grown approximately
1.35 times as much. I think it is not a problem looking
at the attendance growth of pages and understand that
it is almost on all pages, but first of all – on the
pages “Horizont” (an article), “Info_n_News” (some information
and news) and “Main” (the main). For the article and
the main page – the growth is 1.5 times as much (that
is above the average level) and for the page with information
and news - the growth is approximately 2 times as much
(!). But also you can see that one of pages (“Participants”,
participants of studio) - not only does not have growth
but even the attendance falling is fixed.
About what does the obtained statistics speak? To make
a conclusion - it is necessary to know what information
is placed on a site pages. And seriously to think about
the meaning of increase or falling of visitor interests
to either one or other pages.
About my site: seriously having considered the obtained
parameters, I have made the following conclusion: visitors
who repeatedly visit the site form more and more percent
of people. Thus, many of them visit just to find out
news and new information about NooNet. On other pages
there is located mainly seldom-updated information,
which visitors do not want to re-read every time when
they visit the site. As to significant growth of the
article visits - most likely the overwhelming majority
of visitors, who have ignored it for the first time
being on the site, but they have decided to fill this
blank in the following visits.
Conclusions, which allow you to make the described
method of the analysis of the RCounter, are one of the
most important. They allow simultaneously receiving
the advanced performances both about interests of your
audience and about its actions. But, pay attention that
it requires very attentive analysis, in comparison with
results obtained from conclusions ¹5, 15.
General rule: It is not always (not each
month and not for each site), using the described method
of judgements, to receive so correct and useful conclusions.
But from time to time this technique gives very important
results, confirming, or considerably correcting the
usual performances about site audience, its interests,
and actions. The experience shows that this conclusion
gives more interesting results, when the general growth
or the site attendance falling exceeds 30 %. And the
more degree of growth/falling - the above probability
that the analysis of attendance growth/falling for separate
pages will give an interesting result.
Be careful with conclusions: In this
rubric it is necessary again to say already many times
said caution that the above the analysis reliability,
the more “statistical” parameters. And one more caution
– about the conclusion ¹15 - it is not necessary to
use this analysis earlier 10-th of current month, because
it has not enough time yet to get qualitative statistics
for this time.
Conclusion 17. Permitting to evaluate all way of visitors inside your site
Probably, it is the most difficult and ambiguous conclusion,
which can be made from the reports of RCounter (first
version) - about all way of visitors inside your site
(from what to what pages visitors go). Nevertheless
all difficulties and polysemantic interpretations, this
conclusion is of great value and it has a sense to tell
about it below.
To find out from what pages to what ones and how often
transitions happen – first of all it is necessary to
know very well your site structure and secondly, attentively,
look on calendar statistics (better for some months).
The principle of this conclusion - if it is known that
there are links to pages “B”, “C”, “D” only on the page
“A” then, observing the following distribution of hits:
A = 10, B = 1, C = 1, D = 4, it is possible to make
the conclusion: each tenth visitor goes from “A” to
“B”, similarly from “A” to “C” and from “A” to “D” –
here almost each second visitor goes. So, it means that
the link to the page “D” has the greatest popularity.
On the other hand, almost each second visitor from the
page “A” does not go anywhere – simply he/she leaves
the site.
But, I say it again - this conclusion is ambiguous.
The fact is that: 1- it is always possible to fall into
pages “B”, “C”, “D” without visiting “A”, if you use
search systems; 2 - it is possible to call at once all
three pages “B”, “C”, “D” from the page “A” if you “open
links in a new window”. Besides, it is very often to
get on the same site page by different ways.
Exactly, this is the advantage of calendar statistics,
which allows you to minimize ambiguity observing the
attendance for different days. So, if there are no visits
on the page “A” someday and but there are visits on
“B”, “C”, “D” – it means that visitors, really, go to
them using alternative ways. So, we can approximately
evaluate their level of attendance “passing A”. On the
other hand, if there is a link from the page “G” on
the page “D”, then it is possible separately to conduct
observations for the pair “G-D”, it will allow you approximately
to get the popularity of this transition.
It is necessary once again to say: this method of argumentations
does not allow you to determine the popularity of any
transition. It also does not allow you unambiguously
to research the visitor traffic of your site. But, sometimes,
when you know the site structure and 5-10 minutes of
close observation for the calendar statistics, it is
possible to receive an interesting conclusion. Or, in
the worse case, you can understand that it is impossible
to make such conclusion.
General rule: the described method often
is useful for learning such sections of sites as “publications”,
“picture galleries”, “catalog of products ”. There is
often the “main” page in similar sections, from which
(only from it) it is possible to get on other pages
of the section. So, you should only to evaluate “background
of search systems” and the conclusion is ready.
Be careful with conclusions: I just repeat
the remark that this conclusion gives ambiguous results
and it is impossible to give the strict recipe, when
they are authentic or not. It is recommended to use
the conclusions obtained on the basis of the argumentations,
extremely as a “information for reflection ”, but not
as a “guide to action”.
Some other conclusions from the calendar statistics
Generally speaking, looking at the calendar statistics,
it is possible to make some other conclusions, reasoning
analogically, as with the analysis procedure of some
other kinds of statistics considered in the previous
chapters. In some cases the calendar statistics is more
convenient and evident than other kinds of the reports
of the RCounter system, in others - simply supplements
them. I, without long explanations, can specify on these
possibilities and you, I hope, can independently evaluate
their value and convenience in your own case.
- It is possible to analyze the attendance actions
of a separate page, to consider its peaks and recessions
of attendance. If this page itself extremely visited,
it can be possible to make quite reasonable conclusions
on the basis of this statistics;
- Observing “empty” lines (the data lines by pages
which nobody visited for this month, remain empty,
without numbers) you can ask yourself: “Is this page
accessible or may be the link to this page is broken?”;
- Comparing the data in columns “Max” and “Average”,
it is possible to evaluate easily “steepness” of attendance
peaks. When the data in these columns differ no more
than in 1.5 times – it means that there were no big
peaks for this month, the attendance was equal enough.
When the difference is 3 times and more – it means
that there were significant surges of the attendance;
- Considering a parameter in the column “Total” concerning
each page, you can ask yourself: How rationally does
updating of either one or another site section make?
It will be logical to suppose that it is useless to
spend lots of forces for updating poorly visited pages
- it is better to waste your time on development of
those pages, which are popularly. On the other hand,
it is possible to think about how to improve those
poorly popular pages to make them better. With the
help of the calendar statistics such argumentations
is easy to create more correctly than considering
the general level of the page attendance for all time
of site existence. In particular, any page could recently
have a lot of visits and it was a sense to work with
it and for last months the interest to this page has
disappeared and it is possible that there is no sense
to improve it further..
6. Final conclusion
Dear reader, after finishing this document, you probably
have noticed two things: firstly, disadvantages of the
“RCounter” system, when some valuable information simply
does not suffice in the reports and some information
you should get with the help of a calculator. Secondly,
a set of “but”, “probably”, “as a rule” and other similar
turns of speech forcing to doubt in reliability of many
conclusions, which I have described.
I’ll willingly be explained about each of these defects.
As to defects and limitation of the “RCounter” system
– only the version 1.0 of Rcounter had been used for
preparation of this document. But already in that time
I had the version 1.1, which new possibilities permanently
forced me to think “Does it make a sense to write this
document now or it is better to wait for the new version
issue in which many conclusions will be possible to
make without a calculator and additional thoughts?”
But I always responded myself: “It’s worthy of it!”.
Because, I can responsibly declare that users could
not see the version 1.1 …… because the version 2.0 will
be issued. For last months, as well as during preparation
of this paper, we have collected so many
ideas, wishes and lots of improvements (some part from
which, by the way, has already realized) that the following
version of the RCounter system will represent as a conceptually
new product. And I am completely sure that the majority
of pretensions to the contents of the statistical accounts,
which can appear during reading this document, will
be removed once and for all.
As for numerous “but”, “probably” and “as a rule” meeting
in my document. Partly, they are also from imperfection
of the RCounter reports and lack of the information
into them. But there are also other reasons. Among them
the main is my desire to save you from too hasty conclusions,
my attempt to push you to independent argumentations,
in which you can get more high-quality results, better
than those which can be introduced by simple usage of
a calculator and my ready-used advices, which I could
give you instead of my numerous “but”, “probably” and
“as a rule” turns of speech. But there is another reason
- I am not sure completely as far as it is lawful to
transfer the experience of my researches to other sites
in some argumentations. In spite of the fact I have
used Rcounter in different situations and for different
sites with assorted levels of attendance, I think it
is not enough. So, it would be a great idea to get from
you any comments, remarks, additions as well as your
criticism.
Here, I am going to finish this
paper. Let me take leave and if you have any comments,
please write at amfora@lvs.ru.
Good luck on the Net!
7.
Other useful documentation, Internet-resources, addresses
Please, visit www.rcounter.noonet.ru
to find out other documentation and articles about the
RCounter product, new product versions, add-ons, and
plug-ins as well as free version of the product and
some examples of its usage.
Some Internet-resources:
- www.rcounter.noonet.ru
- Internet-site of the RCounter product;
- www.noonet.ru
& www.noonet.ru/eng/
- NooNet Internet-studio (our web-hosting provider,
thank them for a lot of help and useful consulting).
Electronic addresses:
Maxim Bendersky – remlo@noonet.ru.
Alexander Rusin - amfora@lvs.ru;
NooNet studio -
noonet@noonet.ru
|