Commentary

The PPC Controversy Around Credit For Web Site Visits

David Szetela, Clix Marketing

The flawed saying "if you can't measure it, it must not exist" took new meaning when Google began to request budget credit for visits on sites where AdWords serves up ads across the content network.

Now paid search and search engine optimization professionals have begun to pay more attention to the marketing funnel, too. Some urge clients when allocating ad budgets to consider pay-per-click (PPC) ads and query ranking much more than the last click.

Google late last year added a metric in AdWords called view-through conversions that, when pared with multiple exposure attribution, creates "a bit of controversy," though Clix Marketing Founder David Szetela believes it shines light on an important topic.

But while some advertisers want to know about prior site visits because eventually it could lead the visitor to make a purchase, others believe those site visits don't happen enough to consider them in the equation.

Even so, Szetela says traditional analytics don't measure a person's previous visits to a Web site, but marketers need to consider all contributions that lead to conversions. He points to actions prompted by paid search ads or SEO strategies resulting in the sale.

Google knows when someone visits a content site that contains ads it serves up, because the search engine drops a cookie in the browser as the Web site page loads. Szetela notes that if cookies aren't cleared from the browser and a week later the same person goes back to the Web site where ads appeared, clicks on a product and buys it, Google says AdWords deserves credit because it displayed an ad that influenced the consumer to buy the product.

Szetela, whose book (Pay-Per-Click Search Engine Marketing: An Hour a Day) scheduled for release soon, will highlight the strategy. He describes Google's formula like this: one view-through conversion equals one person who loads an ad page, doesn't click on the ad, but later found the ad on the advertiser's site through some other means, and completes the sale.

Should advertisers and marketers give credit to the site visit for the ultimate sale, or factor the site into the cost for conversion?

AdWords can become greedy and count the conversion if the person clicked on an ad at any stage in the conversion process, says Will Critchlow, cofounder of U.K.-based SEO firm Distilled.

"Some systems work on first touch, or 'greedy' attribution, whereby the conversion is attributed to the first way I heard about a site," Critchlow says. "This seems fairer, except it means you would never invest in tools like email marketing that are excellent at getting people to actually buy things."

There are several schools of though, Critchlow says. He describes most sophisticated models assigning some portion of the conversion to each of the touch points along the way. In other words, if someone visits the Web site five times before buying -- three times from different searches, once off a direct visit and once from an email campaign -- you might have a model that assigns 50% to the first touch, 30% to the last touch and a small portion of the remainder to the other touches.

Very few people actually do this, Critchlow says. "I believe some of the big brands work on sophisticated models to understand which bits of the funnel drive conversions," he says. "The whole thing becomes religious when you realize that you are talking about large budgets, and a switch from last touch to first touch attribution, for example, could mean a massive amount of money moving from one department's budget to another department's."

Critchlow says view-through conversions add a whole other layer of complexity, because the person may not actually have seen the advertisement. It could have been below the fold, for example. If a company advertises across many Web sites, the chance that someone will buy the product without seeing the ad remains high.

3 comments about "The PPC Controversy Around Credit For Web Site Visits ".
Check to receive email when comments are posted.
  1. Mark Simmons from marCis interactive, March 3, 2010 at 4:38 p.m.

    This is why you need analytics so that you can segment out the conversions by first and last click. Looking at the user behavior for these two segments will shed light on your overall strategy in this channel. You may find interesting intra-channel trends that affect future efforts.

  2. Fionn Hyndman, March 3, 2010 at 8:32 p.m.

    it's not about the first click or the last click but about the entire 'conversation' with the consumer. If you are really looking at the impact of different media channels you should be able to identify cross media impact to identify the best media mix to acquire consumers, and not just any consumers but the most profitable ones.

    I'm not sure this really rates as a controversy and for people have been looking at multi-channel marketers it has been a question for some time.

    Currently a lot of advertisers use different technology systems to track different ad formats including specialist systems (an ad-server for display, search management for PPC, affiliate system/network for affiliate etc) and through these systems they recognise 1 'sale' or 'action' multiple times and don't have a full understanding of what media 'started' a consumers interest and which 'finished' it (first and last click). One of the biggest things to do initially is make sure advertisers are de-duping the results and then looking at the impact of media channels on the end result.

    If you also look at where the consumer lives (when they have signed up with you) and where they applied from (IP targeting from office etc) you can also infer the impact of offline advertising on the entire journey. That brings offline 'view through' into the picture and although there is no cookie associated it should be taken into account.

  3. Adam Goldberg from ClearSaleing, March 4, 2010 at 11:20 a.m.

    Good blog David.

    Once you start tracking beyond the last click and include all forms of online media including view through impressions you begin to see a much different story as to how customers decide to convert. We classify ads into three buckets to better understand how our clients are reaching the customer throughout the conversation. The buckets we use are Introducer (first step in the Purchase Path) Closer (last step in the Purchase Path) and Influencer (any step(s) in between the Introducer and Closer.

    Typically when we bring a new client onboard 90% of their ads are of the Closer variety which makes sense because they have been using 'Last Click' attribution and therefore only ads that typically perform as Closers get to live on. Because of this we have to work with our clients to add or turn back on more top of the funnel types of ads such as general keywords, display, content networks, and some types of email.

    We then measure the impact of these Introducers and Influencer according the most important metric in business; Profit. As Will Critchlow discussed we use a weighted system that can differ by client to assign profit to all of the ads that occurred in the conversation. To know if this is working we again refer to Profit. If our client's profit is rising then we know the attribution model is effective.

    I do agree that Google is right to get some credit for view-throughs that they track as being in a Purchase Path. However, we find that view-throughs aren't always deserving of credit and other times the amount of credit they receive should not be equal to an add that gets clicked upon.

Next story loading loading..