Have you ever visited a travel website multiple times before booking your trip? Or a visited a retail site multiple times before purchase? How about an automotive site – visiting many times before requesting a brochure or test drive? If so, you’re not alone. In fact, the vast majority of Web users engage in some form of multi-visit or multi-session activity before conversion. This type of behavior is often referred to as deferred, delayed, or latent response. And because it’s happening right now, on your site, it’s critical to understand how it impacts your marketing campaigns, conversion rates, ROI, web analytics, and business as a whole.

From early on, many web analytics and online marketing measurement platforms simply tracked same-session conversion. For example, if a visitor clicked on a Google keyword, came to Omniture.com and submitted a lead in that visit, it would be recorded as a conversion. Fair enough.

But if this visitor visited several times before filling out a lead form, the original keyword would receive absolutely no credit. For marketers, this is a really big deal. In my experience, I’ve seen anywhere from 10%-90% of conversions attributable to multi-session activity.

So as a marketer, you could be understating your campaign conversion by 10-90%! What appears to be your worst performing keyword could actually be your best…poorly performing emails could be driving significant ROI right under your nose…that banner ad on Yahoo that you just killed, it could have been your best acquisition asset.

Fortunately, there are several ways to get your arms around latent response with web analytics and Omniture SiteCatalyst. First, you should scope the issue. Is this really a problem? To answer this question, start with a Visit Number report. This report can be tied to any success event, such as revenues, leads, page views, etc. Look at how many visits it takes before visitors convert. What % of visitors are multi-session? Again, I’ve seen this anywhere from 10%-90%. What is your %? Bookmark this report, because it will provide an important baseline.

Now take a look at your marketing campaigns as measured in Omniture SiteCatalyst. For example, search keywords, email, banner, offline, etc. Ask your SiteCatalyst administrator (if it’s not you) when these different campaigns “expire”. In case you can’t find this, it is under Settings -> Edit Your Commerce Variables.

When does your campaign expire? If it expires on the “Visit”, you’re effectively missing the x% of conversion that you just observed in the Visit Number report. Now you see why latent response is such a big deal!

But wait – before you change anything – talk this over with your team. There may be a good reason you are expiring campaigns on the “Visit”. For example, if you are measuring internal site promotions it often makes more sense to expire on the “Visit” because internal promotions tend to be used as navigation elements more than actual multi-session campaigns.

If there is not a strong business case for keeping the status quo, then definitely change this to reflect multi-session activity. And when you do, I’d recommend you use the SiteCatalyst calendaring event feature to signify that you’ve made this change – because you’ll surely see a substantial uptick in conversion tied to your marketing initiatives and everyone will want to know what happened and who should get the credit! If you’d like to discuss this further, please do not hesitate to email me.

Happy marketing!


Hi Matt, You've hit the nail on the head with this post. What I like about SiteCatalyst is that over time I can measure what the response rate is for different types of advertising, then use that historical data to estimate what the immediate ROAS will be (i.e. within 30 days or so), and what the expected lifetime revenue may be. Although most campaigns will fizzle within the 30 days after advertising has stopped, I often observe latent revenue for campaigns that have been abandoned more than 9 months ago!