At nearly every SES conference, I hear it mentioned that analytics must track 60-90 days out in order to maintain a close watch on latent visits and conversions. In a meeting I had today with a large non-profit based website, the issue of latent tracking over the longer term came up... My question is - is it valuable to track via IP address over the long haul (100-300 days) or are the changing IPs of many ISPs making that data largely irrelevant?
Does anyone here track over the very long term and what results have you seen? Are there significant percentages of visitors that do return for a conversion or action of any kind after 3-4 months?
Great comments. It has been suggested to us we go with cookies for analysis of multiple visits. Most of our conversions come from multiple visits. This thread has convinced us to go wtih cookies.
We directly speak with every sale. It would be very helpful on the web to identify common actions by conversions and currently we are doing this with less than half the conversions.
Because of our direct interaction we understand a lot of the reasons why it often takes people a long time to convert.
Frankly the better we understand all this the better we may be able to either speed up the process or keep more people interested.
Dave
In answer to the original question, I would say that it is not valuable at all to track visitors based on IP address. As has already been mentioned, the best method today for sessionizing visitors is a first party cookie.
The value of tracking visitors across time are many. Sales cycle measurement, latency, recency, lifetime value, etc. . .
Not every analytics tool allows for tracking a visitor over time but Google Analytics has some reports for this and pretty much all the top tier analytics tools (Coremetrics, Omniture, WebTrends Web Side Story) have this capability, although we should not forget Matt Belkin's blog posts trying to diminish the importance of visitor tracking.
Rand, I have a lot more to say on this topic but don't have time to write it here. Please feel free to contact me directly I would be happy to have a more in depth conversation with you on this topic.
The bad part about Google analytics is that it can't track Variables. So that report that shows "Site Overlay" our action buttons that have unique variable are being ignored.
I think Google analytics should be used as a secondary tool, OR the only tool for a company that can't spend $200/month.
I am not totally sure what you meant by variables. But I think you meant query parameters in the URL. And in that case Google Analytics does record that. You might have to fiddle with the settings.
IP addresses aren't reliable. Certain types of IP addresses are much less likely to change. My IP address at home hasn't changed in months. Dial-up users get a new IP address every time, but with proxies they may appear to have the same IP address @ every visit, as will many other customers of the same ISP. AOL users will hit you through one of many proxy servers.
When I was moderating at the IHY forums, checking IP addresses on "unusual" posts often made us wonder if certain groups of members were actually the same person, but you can't know for sure. Many of the UK based members came in through the same single IP address, which was a proxy server for an ISP over there.
If you want to track long term visitor behavior, I think you have to use cookies. The problem with long term cookie tracking is that you usually can't measure/estimate how many drop out of the tracking system.
From forums and other sources where we actually can tie a cookie to an individual user, and track the individuals' visits, I'd say in the timeframe you're talking about, somewhere between 20-50% will dump their cookies.
Dan,
Jupiter Research published a report that indicated that the deletion rate for third party cookies was somewhere in the neighborhood of 30% so your numbers are pretty accurate. However, the use of a first party cookie will cause that number to drop to less than 10% in most cases.
A typical user will delete cookies because a spyware removal tool told them to, and not because they went to Tools>internet option>delete cookies. Spyware removal tools typically target third party cookies.
In case a definition is needed: first party cookie = cookie served from the domain the visitor is currently on Third party cookie = cookie served from a different domain than the visitor is on
My comcast IP address hasn't changed for a long time, but it will if I disconnect for 3 days (says the rep). so us online addicts probably have much more stable IP addresses than the general population
Another alternative to various Tracking Visitor Conversions tools available on the net is mvispy.com I have been using the service for past few months and it has worked out well for me. I have been able to do live user tracking. This is an excerpt from their website.Url: https://www.mvispy.comContent: Tracking Visitor Conversions Paths MVI software enables site owners to set conversion alerts when Visitors view certain destination pages, by the amount of time, completing forms or submit orders. These alerts trigger the entry of that visitor data into conversion table to be viewed on MVI SPY or the data can be sent automatically to a Sales Lead software system, this data clearly defines each successful visit. Simply put, MVI SPY tracks all conversions generated from your online web site capturing names, addresses and pertinent data and ties information directly to the exact keyword or ad source. MVI SPY is an award winning, online-to-offline ad tracking solution that gives you unprecedented visibility into your entire conversion performance by tracking all visitor views and conversions generated from by your site or online advertising.Research indicates that an astounding 60 percent of searchers complete their purchase offline following their search activity. Conversion tracking tools measure online conversions, by form submits, registration, free downloads, newsletter sign ups, and also measures the conversion by preset paths followed. In addition to tracking your online conversions, MVI SPY inserts conversions directly into database for sales staff use by keyword, paths followed, or forms completed and keywords used.
So to that end..when setting cookie for tracking purposes, what should the cookie contain? the current IP, some random number? We use LiveStats.XSP 8 and there is a capability to view visitors by cookie, but the documentation is scarce.
Does Google Analytics offer latent conversion tracking?
I looked at Google Analytics when it first came out and there was some limited converstion tracking and on report did talk about returning visitors, but from memory that's all I remember.
Google Analytics does set a cookie to track returning visitors (instead of IP address), so the answer is they can. Anyone actually using Google Analytics?
I manage several web analytic products for several clients all for the purpose of eveluating the success of their SEO, SEM and other on-line marketing activities.
In all instances we don't track by IP simply because they do change and with the use of proxy servers, thousands of people could have the same IP address. Instead we use persisent cookies. Cookies do have their limits (about 5-10% of people don't accept them from what my reports tell me).
The huge benefit is, yes traffic is significant.
A visitor my download the product brochure today (first visit0, and come back next week or even a month from now to purchase the product. How many times did they visit a site before they made a purchase? These are important questions that need to be answered when evaluating user behavoir and improving a sites content and usability.
What about repeat visitors and their behavoir.
I believe it is critical to track visitors over the long haul.
As to software, my favourite is WebTrends for easy of use in reading the reports. However all the major products on the market these days offer very similar reporting. So you basically need to try out the free trials to see which one you like.
My question is - is it valuable to track via IP address over the long haul (100-300 days) or are the changing IPs of many ISPs making that data largely irrelevant?
It really depends on questions like
* who is your audience? Where are the users coming from (personal or corporate users)?
This has huge affect on IP-based approach. Most company users have/use fixed IP's or they become trough proxy/nat; home users have varying ip's or they use isp proxy/nat.
* how large site are we dealing with? As irrelevant as it may seem, this is really closely tied to above question. The larger the site, the more likely it is to attract versatile traffic.
* how long a period? The longer the time, the more likely you end up having users with varying ip's.
* What are you actually trying to track? Unique visitors, business names or what?
Based on my own experience I'd say the skew is somewhere between 10-30% for 100 day tracking, and 15-40% for 300 day tracking. So I would call this a very inaccurate way for tracking unique visitors as a standalone approach; when you add cookies and/or other tracking data/methods you end up having much more valid data.
It is a good idea to track an IP over the long haul because you want to know if there are any repeat buyers. It is not always 100% accurate, but it is still worth looking at.
A simple solution that I came up with to deal with that was just to use a cookie.
If the cookie was not set on visit, then set the cookie with the client IP address.
If the cookie is set then use the stored cookie to set th e IP address to be recorded by the logging software.
In this way it does not matter that the IP address is changing - you always log be the original IP address.
What are some good, no, rather great software to track visitors for such long periods? I have been thinking of writing it myself, and probably will. But it cant hurt to look at options, or atleast get some ideas for features :)
Tracking over a long period is very interesting for one of the sites I manage (insurance). We track quotes through the whole cycle and can see the referrer details for each quote and its status.
Some people take a few weeks to buy and we often see days where we have given 100 quotes for example and had 110 sales as people return to pay for their policy.
Mad4:
Do you interact on the phone or in person with your customers?
I am sure that this varies by industry and product... however the long term conversion stats (not necessarily by IP) are important to affiliates who might have 30 day or 60 day cookies.
Rand ... THANK YOU for starting this topic.
This is a huge problem for our company. 80% of our sales come from "latent conversion" (2nd, 3rd... visits). It's very hard to track these.
I use Indextools, and after speaking with these guys multiple times, the only improvement they made was that you could see IPs of last visitors 500 per page.
What I do is, I try to track the first visit by IP address but it's very hard to do. AOL IP addresses do change and many visits are not tracked.
Out of 100% sales, I am able to track down about 40% of them, thus my numbers are not as accurate as I want them to be (especially since we don't do a huge volume of sales).
I spoke to other tracking companies (Webtrends, ClickTracks...) about this and some of them had the tool for latent conversion ... nevertheless, it was a bit confusing to use (or maybe their presentation was confusing).
I am still trying to battle this problem. Rand ... what do you got on this subject? ;-)
Rand:
I check conversions every month. #'s of visits before a conversion ranges from 1 to over 20 visits. Average number of visits typically ranges between 2.5-3 visits for a conversion.
As an old line brick and mortar business we have known this for years and it shows up on the web. Some of our customers don't purchase our services for a long time.
As a person trained in sales I always liked the idea of a difference between a "big sale" and a "small quick sale".
It doesn't take me any time to buy batteries, milk, gas, or spend for a nice dinner for an excellent evening.
If I'm buying a house or a car, looking at buying significant services for my business, or making a big change I could spend a lot of time on the decision.
Big sales typically take more consultative sales effort. Likewise customers making "big" decisions can take considerable time before they purchase.
Tracking over time is possibly or probably a function of the type of decision. If I'm selling shoes, coats, video equipment on the web I wouldn't worry about it. If I'm selling bigger services on the web it could be well worth it.
Dave
Buying habits haven't changed.