Pear Analytics

Every now and then we have customers who want to cancel their service because the SEO effort has not generated an ROI in terms of sales through the website.  While this is certainly reasonable to assume, it’s almost an unfair proposition given the amount of external factors unrelated to SEO that can drive a “sale” on a website.

Consider that the first and foremost goal of search engine optimization is to raise the visibility of the website in the search engines, by increasing the organic rankings for a variety of keywords related to the business.

So the basic equation then, is that SEO leads to higher rankings, higher rankings lead to an increase in traffic, and the increase in traffic leads to an increase in sales.  Seems reasonable on the surface, right?

However, as we analyze this thought process, we see that the “increase in sales” goal, or directive,  is all the way at the end of the funnel.  When most SEO engagements start off, the conversation revolves around keywords of importance, content, links and maybe expected traffic increases — but almost never conversion rates.  You might be able to fairly assume that the existing conversion rate could be applied to the new traffic generated from the SEO effort, but the discussion really requires further investigation of the design, usability, and other factors that lead to a sale.  Now we’re not just talking about SEO; we’re also talking about conversion optimization, and A/B testing to get to the desired results–a service that the SEO firm may not be performing, involved with, or even have control over.

The second problem lies in the “keywords related to the business” statement.  I can tell you that most customers don’t know for certain which keywords drive sales.  If we did, we’d just focus on only those words and solve the sales problem.  But the answer is far more convoluted than you would think.  They may know what they’d like to be found for, but for the most part, those terms are not keywords with “buying intent.”  Also, what about click attribution?  When does PPC get the credit, versus SEO?

So I offer an attempt at a solution to this madness.

First, with customers who are “green” at the whole SEO or Internet marketing thing, you could propose a 3 or 4 month PPC campaign, target a variety of keywords, and measure which had the best conversion rate.  You could take this same performance logic and apply it to the SEO campaign, and hope for the same, or similar results.  In my opinion, visits through a paid channel should come with increased buying intent, as opposed to the organic channel, which could be more research-oriented visits — but I can’t say for sure.

Second, we could perhaps do a better job as SEO-ers in terms of classifying the keyword targets into the marketing/sales funnel.  The typical marketing or sales funnel has 3 sections:  Awareness, Consideration and Purchase.  Ideally, a keyword in the Awareness channel will create a lead for you to nurture, but not a hard sale, as might a keyword in the Purchase funnel.  In my illustration of a fictitious hosting company, you can see how you might bucket these keywords:

So as we get better as marketers about driving results, it’s clear that the most important step is to set the expectations correctly up front — not just in terms of rankings or traffic, but also about sales. Have that conversation up front and make sure everyone knows what they’re signing up for.  Because if you’re used to a sprint, you’ll need to start training for a marathon.

2 Comments »Tags: first click attribution, marketing funnel, seo efforts

As marketers, we would love to know how all of our leads originated.  Far too often, PPC gets the boot with poor cost per lead (CPL) figures, and can be mostly blamed on attribution issues.  Attribution is simply defined as assigning credit to the source which generated the initial (or final) action.  So for example, let’s say you’re running a paid search campaign and you get 100 new visitors to your website.  However, only 2 of them make a purchase.  Unbeknownst to you, an additional 5 purchases came in over the next 7-10 days, but the source was either “direct” or from some organic search term.  In this case, the marketer might attribute 2 sales to PPC, and 5 to SEO, and SEO might win in terms of lower cost per lead.  So what’s the solution?

Multi-Channel Analytics & Funnel Analysis

Google recently launched a new feature which attempts to solve this attribution problem, and show improved ROI on AdWords.  Of course, Google is always interested in finding ways to show PPC really does work, and in this case, rightly so.  But check this out.  If you’ve properly tied your AdWords account to your GA account, and setup goals, you will see this Venn diagram.  You’ll find it under the Conversions section.

In this small e-commerce website, we can see that 4.65% of the conversions occurred when a visitor clicked on a paid ad, and THEN came in later by typing in the domain name directly.  Without this, the marketer may be quick to assume that the 2 sales from the direct path were not in any way influenced by paid search.  Now, this is neat.  I can also see what my “lag time” is between first click attribution, and the actual point of purchase.  Why do you care about this?  Well, if you have enough sales happening too far after first click, you may want to try to run a promotion to those visitors who are “on the fence” about your product or service.

First Click Attribution vs. Last Click Attribution

Marketers often debate whether to give the “first click” credit for the sale, or the “last click”.  Some might argue that even though the PPC ad created the interest and got the visitor into the funnel, it was the “last click” that finally sold them, and so that should get the credit.  This is certainly a valid point, but you have to look at a few more things, such as: a) was there a coupon or discount that may have led to the final decision?; b) was the PPC landing page set up as a lead generation, or a hard sale (i.e. “Buy Now”)?; c) was there a pricing matrix or downloadable whitepaper that may have led to the purchase decision?  In other words, did the last click lead to a decision-making point?  If so, maybe then you assign the credit to the last click, and measure PPC in terms of “cost per lead” only.

1 Comment »Tags: first click attribution, last click attribution, multi-channel analytics