Yes, I realize that measuring SEO ROI sounds simple enough. Its not so much a case of measuring per se, as a case of is what youre measuring enough?

Case in point: In-house SEO:
Tasked with measuring and comparing last years SEO ROI to this years, its a given that one of the basic formulae below will be used:

  • SEO$ROI = (Profit-Investment)/Investment
  • Simple Profit = (Revenue-Cost)

    But are either actually a true reflection of the returns from SEO? No. For three primary reasons:

    1. SEO: Supportive Cannibal - Quantitative
    Scenario
    People are exposed to brand or emotive messaging telling them they need a product or service type constantly. Think TV, radio, billboards, moms friend, wifes day care owner, the neighbours, banner ads, email, signage, packaging, shop windows, text messages, forums, blogs We are inundated with thousands of marketing messages (ad clutter) every day.

    Process
    People remember snippets of things important or of interest to them, and many go home and search for a brand name, a tag-line, or an advertised keyphrase.

    Result
    Offline exposure is far harder to include in an attribution model than other online initiatives, but regardless SEO is cannibalizing - to an extent - the revenues that should at the very least, in part, be attributed to other advertising channels. SEO is due a portion of that attribution due to the fact that if SEO wasnt doing its job, those brand name listings ranking so well might not be, and those generic non-brand terms for which your site ranks well " wouldnt be.

    2. SEO: Ignored Facilitator - Quantitative
    Scenario
    The evil twin sister of the supportive cannibal; as an ignored facilitator, SEO contributes to a host of additional online revenues that are simply not attributed to SEO efforts.

    Consider
    SEO optimizes a PR release for online publication; adding relevant anchor text, optimizing the title and checking content for relevance, etc. Apart from normal releases (now optimized by SEO), SEO also blasts it through PRWeb (for example). Are you measuring referral traffic increase from PRWeb along with conversion from PRWeb? Many dont and miss the SEO contribution to conversions.

    What can you do?
    Use a dedicated landing page on a unique URL and track referral traffic and conversions by referring site for these releases " you may be quite surprised at what SEO achieves. Also, consider running 2 releases that are similar in target readership, one with SEO and one without, and check SERP exposure (PRWeb is great with that) as well as referral traffic and conversions and conversion path.

    Consider
    Your SEO is probably encouraging backlinking of one sort or another through the socnets (Social Networks), integrating with social marketing and through PR integrating with..well..PR, relevant directory listings and a host of other methods. Are you measuring the impact of additional referral traffic on your conversions in your analytics? If so, are you fairly giving SEO some well deserved kudos?

    What can you do?
    Request your SEO to take an analytics grab of traffic in numbers as well as in conversions for referral traffic prior to a socnet initiative or directory drive (etc).

    Im not suggesting rushing for backlinks from anywhere in a hurry " Im saying, if no backlinking or socnet initiatives with an SEO flavour were undertaken last year, then measure 6 months of last year comparable to 6 months of this year that you DO employ those 2 strategies on a continuous basis, and take into consideration share of overall traffic as well as share of overall $Y (income/revenue), not just absolute data. Again, you may be surprised.

    3. SEO is the baby thrown out with the bath-water - qualitative
    The impact of SEO from a brand exposure on the search engines is nearly always ignored when it comes to the long-tail or seemingly irrelevant search phrases that end up in a conversion.

    Consider
    If SEO is consistently ranking your product or service for a range of relevant non-brand/generic terms, people may come back via search and swap as well as search and time-out behavior. That exposure in the SERPs (search engine results pages) on these terms builds brand awareness in searchers, but may never be attributed to SEO as people exposed to the brand via this method may very often choose to walk into a brick and mortar, or go to E-bay, Amazon, PC-World etc and end up buying from another online retail distributor

    What can you do?
    This is nearly impossible to deal with, but not completely, and warrants an entire post of its own. Please bear it in mind nonetheless

    Conclusion
    If SEO is effective in ranking your site for all brand and most non-brand terms, then SEO deserves the benefit of the doubt in terms of being undervalued in traditional ROI measurement. Attribution analytics models can go part way in leveling the playing field and giving SEO its due, but not all the way. Dont undervalue the power, exposure and contribution to overall conversions of SEO.

    My advice? Give SEO more exposure in-house to mix with all marketing to ensure an holistic approach, enable them, facilitate their engagement and give them the dev, QA resources and deploy/release time to develop dedicated landing pages. Make sure your analytics folks speak to SEO on a regular basis and best case scenario, give your SEO full access to your analytics if you dont already.

    Caveat: In-house has already hired SEOs based on the value of their CVs, experience and successes; for agencies or outsourcing, ensure you are entirely comfortable with the NDA in this particular sphere before sharing all analytics and strategy. If you want effective SEO across every facet of your online presence, they need to be involved from the IA stage of each new strategy, and they should be required to contribute. You have the final decision always, but you may benefit enormously if you listen to what they have to say.

    Laura is Senior editor and a blogger at SEMInsights.com. She is also Web Marketing Manager - SEO for Intuit Canada. Nothing she says here may or should be attached to any employer, current or previous. The views expressed on this page are Lauras alone and not those of her employer. Follow Laura on Twitter.