Question

Topic: SEO/SEM

Marketingprofs: How Do You Prove Roi From Seo?

Posted by Anonymous on 500 Points
We are creating a new site and have been approached to by a company suggesting we pre-build search engine optimization. Our marketing budget demands a high ROI for each activity – how can we justify the expenditure on SEO if the service provider cannot demonstrate ROI? And yes, we have an advanced tracking system, but apparently that is not enough to identify what revenue is generated by SEO.
To continue reading this question and the solution, sign up ... it's free!

RESPONSES

  • Posted by Annelies on Member
    There are several ways:
    Google analytics is a good tool for this. It shows you which traffic comes from search engines and you can make a difference between paid and non-paid traffic.

    Another approach is: If you know where all paid traffic is coming from by e.g. putting counters on every link/banner click that is coming in. You can count link this: Free traffic including SEO = All traffic - counted paid traffic. From the free traffic you need to minus the direct traffic and you have an indication on your SEO traffic.

    Good luck!
  • Posted by Lorenz Lammens on Accepted
    Return on investment from SEO cannot be clearly 'proven' in the same way that ROI from a machine can be 'proven.' because:

    the return lasts for an indefinite period and at various levels of value.
    --you do not know how long the increase in traffic will last
    --how much of future traffic can be attributed to your contribution. (immediate traffic increases can be easier to attribute)
    --You do not know the future value of that traffic

    You can return a very useful analysis of this investment by looking at meaningful metrics. You'll need to show context somehow. Either trends or 'before & after photos' is the usual choice. Here are a few:

    --General Metrics
    --unique visits
    --unique organic SE visits
    --Loyal visitors (define this metric somehow and stick to it)
    --Visits resulting in more then X page views (or time on site)
    --Visits displaying a certain level of engagement (you'll probably have to customize this metric if you choose to use it)

    --Direct Outcomes
    --leads
    --sales
    --revenue
    --mailing list growth
    --other goal conversions

    --Branding Metrics
    --Direct Visits
    --Brand specific search terms
    --returning visitors
    --long term loyalty
    --engagement metrics
    --Search terms specific to sub-brands r brand specific product names
    --Identified visitors (signed in)
    --repeat purchasers (or conversions of other types)
    --participation in online activities (forum posts, comments, reviews, send to a friend traffic, wiki's, etc.)

    Naturally, if the site is an e-commerce site with clear, direct goals and most purchases are incidental (visitors search a a term & buy 2 minutes later) the ROI is more of a 'real' provable accounting metric. If your goals or the selling/marketing/informing/educating/persuading process is more complex, ROI is further from accounting ROI as the return is what they would call intangible assets. These latter cases are more difficult to measure or prove but they can usually be demonstrated.

    Which/How many you choose & how you report on it will vary depending on the purpose of a site. If IM is an important component in their marketing, monitoring these will be important anyway.
  • Posted by Harry Hallman on Member
    You did not mention if your new site is an ecommerce site. If it is you should be able to measure where your sales come from, using analytics programs such as Google offers. You can also set up goals to measure specific pages visits and of course, you can measure registrations. It all depends on what your goals are for the site.

    I have seen sites that use outside SEO companies dramatically increase site visits and I have seen it not have much effect. It depends on the people doing the SEO work.

    I always find it interesting that some marketers (not you since ROI is the main goal of your marketing strategy) are very comfortable spending vast amounts on say print ads.
    These marketers have very little idea if these print ads actually create sales or that they are even read. These same marketers have a problem spending a fraction of that amount on SEO or Social Network Marketing, which is more measurable.
  • Posted by Jay Hamilton-Roth on Member
    You can measure ROI after the fact, but you're wanting numbers before you start. I'll echo Mike Ashworth's post: improve your % of your conversions now, then improve the SEO to get more targeted traffic (and therefore more conversions).
  • Posted by Chris Blackman on Member
    John

    This is almost a chicken and egg situation.

    Sounds like you're deadlocked because you've been told not to undertake a project unless the ROI can be demonstrated. But to prove something, you need evidence, which is based on facts, which you don't have.

    Before the fact, any ROI projections will be just that, projections. Estimates. Maybe even guesses. But not facts. The only way you can get ROI facts - concrete evidence, proof - is by completing the project, and conducting post-analysis.

    So what you need to break this nexus is for the service provider to give you some guarantees. And the guarantee has to be meaningful, such as a full refund if the perfromance and ROI hurdles aren't met.

    Otherwise your project approval process will have to accept the risk involved in undertaking a project with potentially large gains but no guarantees.

    It's about balancing the business risk.

    You can mitigate the risk by checking the service provider out carefully. Are their assertions and claims backed up? Have they done similar work for other companies like yours? How did that work out?

    You may gain confidence enough to proceed. But if your management is looking for an ironclad guarantee, you'll have to get that from the service provider, which seems to me to be the only party in a position to make that kind of an offer.

    Hope that helps.

    ChrisB

Post a Comment