Web log analysis
May 14, 2008
I’ve posted various metrics about BookMooch before, using numbers from the BookMooch Statistics page. Today, I decided to take all the monthly web logs, combine them into a big excel file, and see what that looks like.
My logs start November 1st, 2007: before then I wasn’t saving the web site log data. On most of these charts, I’ve added a 14-day trend-line, to quiet the noisy day-to-day fluctuations.
This chart of daily unique visitors (ie, individual people) shows us at around 18,000 daily visitors, which is 48% more than 6 months ago.
This chart of daily web page hits shows a 45% growth in the past 6 months, with 300,000 page hits per day currently.
Both of these charts are understandable when you look at our BookMooch members listing books to give chart, which also shows a 50% growth in the past 6 months.
This chart shows when people visit BookMooch. Predictably, it’s a bell curve, with the bulk between 8am and 8pm (Pacific Time). I made a similar chart for day-of-the-week, but I’m not putting it on the blog, because it showed that usage is almost identical for each day, except that Wednesday was about 20% less busy.
This chart shows how long people stay on the BookMooch web site. Most interesting to me was how many people stay a fairly long time, with 45% of visitors staying more than 15 minutes. That is quite unusual. Note that I removed from this chart people who stay less than 30 seconds, since they obviously didn’t find BM interesting, and left right away — I was interested in seeing how long everyone who actually uses the site stays.
No surprise here, 50% of BM visitors use Internet Explorer, and 37% use Firefox.
When I get started making charts, I have trouble stopping!
Here are two more, and then — really! — I’ll get back to work.
I’ve always wondered what our quarterly growth looks like, since that’s how public companies report their numbers.
This chart shows how much more mooching occurs each quarter vs the previous quarter. Yes, I know that we started at high growth rates (that’s why the curve goes downwards from right to left) as that’s always the case with a new company, since you’re starting out with very small numbers, so it’s easy to have growth.
For the past 9 months, we’ve stabilized at around a 27% quarterly growth in mooching of books:
Interestingly, the growth in active members (people who have books listed for giving) is quite a bit lower than the growth in mooching. For the past 11 months, we’ve been around a 12% quarterly growth rate in active members. This shows that BookMooch is become “more useful”, in the sense that the people who are on the site are mooching more and more, at a much higher growth rate than the acquisition of new members:
BookMooch Blog 
I love charts, too! As a user with a totally non-technical background, it would be nice to better understand exactly what some of these charts are showing. (Sometimes I think techies forget that not all of us speak web-talk fluently.) For example, “unique daily visitor” sounds straight forward. But, is that unique log-ins? Then what about guests who don’t log in? Or based on the computer logged in from? What about when I log in at lunch from the office and again from home after dinner? Similarly, with “hit.” Is every page I look at a hit every time I look at it? Then my habit of jumping around pages pretty randomly then backtracking would count as more hits than if I was more systematic. Just interested.
I’ve been wanting to do some analysis of the XML downloads that are available ever since John put out his call for economists last month, but I ran into a problem in that the file containing user information contains invalid XML characters. This gags all of the import tools I’ve tried using, so I can’t really do much of anything with the data. I suspect there may be some interesting trends that could be teased out of the data based on day-to-day changes, but for that to be practical I need to be able to import the files quickly.
I have just enough OC traits to like tracking stuff, so the charts are interesting.
I am probably one of those who “stay” on BM for 30 minutes or more, because I am wandering around other book sites to compare my wishlist to prices elsewhere, how badly do I want to buy versus mooch a book, etc.
In my “wanderings” I find the books listed by topic section of BM not helpful. The ex-librarian in me cringes badly because books and topic do not match.
Ex: just now, in history, I find listing for the Physician Desk Reference, one for 101 Dalmations, one for A Charlie Brown Christmas,
sigh……
Sure would be faster to browse and to mooch if, say, the history topic section contained only books related to history.
How exactly is the topic determined when a book is listed in an inventory???
Hi! I love the charts. Question though – I tend to leave bookmooch logged in all day while I’m at work, even if I’m not actively doing anything in it. That way, it saves me the trouble of re-logging in if I go back at lunch time and it shows “logged in now” which might mean someone would choose to mooch from me instead of someone else. Anyway, I’m thinking this type of behavior probably messes up your ‘time spent at the site’ chart so I wanted to give you a head’s up in case others are doing it too.
I am also one who stays logged in for extended periods, even if I am not actively using Bookmooch. I am surprised the growth rate is as low as it is. I recommend the site to everyone I know who loves the read and even random people in the bookstore. I agree with Kathy about hoping to increase someone’s chances of mooching from me if I currently show “logged in now.” I personally take it as I am more likely to get a response. I always look at the last time someone was logged in before placing a mooch. Thanks for the charts, John. They were very informative.
I’m one of those who make up the <30s contingent. It’s partly ADHD I suppose:
Click on Bookmooch,
hit return on the popup login window,
click on wishlist,
click on “all”,
pagedown the list to see if anything is available to mooch,
close the browser.
Repeat every 10 minutes. It really does take less than 30 seconds. I suspect there are others like me that screw up your stats.
Is it better to sign out when you’re not using bookmooch? Does one person constantly signed in mean there’s one less open slot for someone else? It used to be that way with online communities: a system could only support x number of users logged in at the same time, and too many inactive users still logged in would bog down the entire system and keep others from being able to sign in.
Coqueline —
Please do not open a second account for angel use, but rather use your original account as others do for this purpose.
We reply to all reports with 12h or less. The problem is likley related to your attempt to open a second account which while needed in some special cases, generally requires admin review first.
Mark W
Bookmooch Team
Is it better to sign out when you’re not using bookmooch? Does one person constantly signed in mean there’s one less open slot for someone else?
It does absolutely no harm to stay logged into BookMooch all day, you’re not taking anyone else’s slot. If that works best for you, go with it!
-john
Hi John, it’s great to have an update on how BookMooch is doing. Thanks for everything you do. It’s nice to see the site is constantly growing.
Just wondering, is revising the current “ask first” setup still on the to-do list?
Debra
Hi
It would be really interesting to see which countries are mooching and then link that to the number of active members- see where the international moochers are
Also would be interesting to see the proportion of moochers providing international mooching to local mooching as well
Doug
Interesting charts, John.
Oh, and the Bookmooch cards that I had asked for arrived today. Thank you!
John,
I know this does not apply to this blog, but since it is still on people’s minds as mentioned above, I have some ideas I just thought I’d share about rectifying the “ask first” problem.
1) Could you make it so that when a person was listed as “ask first”, any international mooch pulled up a window that asked the moocher: ‘Have you emailed the person and received permission to mooch this book?’ If they click yes, the mooch can proceed; if they click no, the mooch is stopped and they are prompted to email the person. If it could be set up this way, it would prevent us making the mistake of mooching without asking. (However, if a person were not honest this would not work…) I imagine a lot of the time a mooch is made w/o asking first is just due to an oversight; it is easy to overlook that the moocher requests to be asked first. I have almost completed a mooch different times without asking first before I realized I was supposed to ask.
2) Is there a way to lock out international moochers for people who want to be asked first, similar to how we cannot mooch from someone international who will only send to their country, with a difference being that we as individuals could go onto BookMooch and unlock a particular book if someone international has requested it and we are able to send it to them?
I just had those ideas so I thought I’d share. I’d be excited if one of these ideas proves helpful.
I’d be interested to see some stats on how many people sign on, list some books (perhaps at least ten), but get no mooches and end up going away pretty quickly. My suspicion is that there are a significant number of them. I also suspect that it would not take a lot to retain a good percentage of them, if they were to get just a mooch or two. If my assumptions are correct, I’d like to see something systematic that aims at retention.
For example, could we have the system track new members who have some minimum number of books in their inventory (again, ten might be a good minimum), and who also do not get any mooches for perhaps their first week. I’m quite sure we could get a goodly number of folks on a list who would volunteer to mooch a book from such folks, and also extend a welcome and offer assistance to such Newbies. Perhaps something like the ransom wishlist process could be implemented to choose a member from that list, and request that they help get the new member started. If they could not do it, another member from the list could be randomly chosen.
My suggestion makes assumptions about the ability to track new members and their inventories, and whether or not they get any mooches, and I really have no idea if these are reasonable assumptions. It also assumes that the numbers of such short-lived members is large enough to warrant the effort to retain them, and even whether or not such information could be available.
What say you John? If John says it’s possible, anyone else have some thoughts about this or other good uses of Web log stats?
Stuart (USA: CA)
(scamp8)
To Rachel:
I think a lot of people who are listed as ‘Ask first’ don’t really WANT to be emailed every single time an international moocher is interested in their books, but simply because the only other options are ‘Worldwide’ and ‘Only to my country’. These people usually write conditions on what and where they’d send internationally on their profile, and it’s pretty clear.
Being locked up from mooching at all until email correspondence is confirmed is basically putting most international moochers at the bottom of the food chain, since they’ll never be fast enough to mooch anything that is remotely high on demand.
If there ever is a new category like ‘Send internationally on conditions’, I would jump ship right away from being an Ask-First-er to the other one.
I got the cards & thanks!
I’ve started including them in jewelers orders & promotional mailers.
My e-commerce customers are also getting a card with their orders.
I travel around the southeast visiting wholesale accounts & will be leaving cards in Books a Million’s & Barnes & Nobles unless you tell me not to!!!
Love your work!!!!!!!
I’m glad this place is still growing 🙂 More people really ought to know about it, especially if they live in a town that doesn’t have a second hand bookstore [anymore].
I qualify as someone who doesn’t have an affordable second-hand bookstore any more.
I do live in a town with a gloriously wonderful library system.
Best,
Evy
http://www.bookmooch.com/m/inventory/evy47
Love the site, the book senders and moochers, the camaraderie, the selection, agree that the topics and books are misfiled-oh well-it is free!!
Very happy to have found you, I tell everyone I know-we’re all readers.
Keep the charts coming, John.
Sincerely,
LMO
[…] the other hand, PaperBackSwap’s traffic is higher than that of BookMooch and they have more books available. They’re also clearly out to be the Oprah of the modern […]