Why Conversions Can Be a Messy SEO Metric

SMS Text

Seems I am still running into SEOs that believe, however misguided, that conversions are somehow their domain. This seems speak to the fact that many search optimizers may not have a grip on the marketing model. While I’m not as well versed as I once was, let’s at least take a stab at this…

For starters there are simply too many potential factors out of the average SEOs control that can affect the over-all conversion rate on a given website/page. I know this is unlikely a popular view, but hear me out for a few moments.

Look! Sheeple


Factors affecting conversion rates

The first thing to consider is the actual elements that affect conversions. So let’s review a few of them shall we? Here are a few elements which can affect conversion rates;

  • Product/Service offering; right off the bat product/service choice and ultimately pricing will play into conversion rates.
  • Website infrastructure; a website that is not intuitive and user friendly will undoubtedly have more sales friction (abandoned carts, missed lead capture).
  • Page Copy; professionalism, unique selling proposition and call to action. Obviously an important consideration in the process. Developing buyer motivation and the moving the buying cycle is massive.
  • Incentives; other products or value-adds for the core offering. Many times one has to offer added value to further cement the final sale ( or upsell/downsell etc..).
  • Continuous improvement process; ongoing A/B, MVT testing and tweaking. This type of page optimization is not traditionally the domain of SEOs.
  • Customer Service; oddly this can have an affect if people are engaging in WOM and thus visitors (even if only doing a navigational brand search) are somewhat pre-qualified when entering the site. Even past (happy) customers will use a navigational search to go back to the site many times.

And that doesn’t even cover all the elements of the consumer purchasing decision process. What I hope it does do, is give you an idea of the myriad of elements that come into play when one is talking conversions. If you advocate a ‘pay for performance’ model for SEO, then do so at your own risk. Here’s an interesting bit of research to give you further insight;

Purchasing decision data


What DO we do?

Most certainly SEO and conversions are related and can play together. As SEOs we need to understand them and tailor targeting strategies accordingly. But our job is to bring ‘targeted’ traffic. This is not the same as being involved in the conversion optimization process. I don’t believe the average SEO should be fooled into believing otherwise.

And so what is it that the average SEO does that plays into the conversions realm? A few that I can think of include;

  • Keyword targeting – obviously one needs a solid KW research program to ensure you are going after the appropriate query spaces for the best possible potential. Ranking for terms that are poorly targeted can obviously bring consumers with a low threshold of converting.
  • Rankings – being above the fold is certainly going to improve the SERP CTR. This will bring greater amounts of targeted traffic. To that end, this IS an area where SEOs can affect the over-all program and conversions.
  • Page TITLE – not only does this come into play as far as ranking signals are concerned, but it can also be an important aspect of the SERP click-through rates once you’ve made it to page 1.
  • Meta-Description/Snippet – also related to the above, though not generally an important ranking factor, a well crafted snippet can play into SERP conversions… (though the engines can also choose page text)
  • Link building – if we want to be ranking for the target terms, in most cases some tightly targeted links will be needed to gain placements most likely to produce the most targeted traffic.

And really, that is about it. Sure, there are many of us that work with actual page content and even call-to-action etc… but that’s no longer SEO now is it? It is a cross-over.

The actual process of SEO doesn’t really touch on the many elements (we’ve outline) that affect conversions. Yes, we are a link in the chain, but we’re not by any means dealing with ALL the areas that come into play.

SEO pay for performance


Don’t shoot the messenger

The next time you hear an SEO talk about conversions or even worse, pay-for-performance, you’d do well to stop and consider this advice. Not only are there some inherent risks in thinking this way, there is a fair bit of naivety as well (not to mention insulting to the entire conversion optimization industry).

Search optimizers rarely have the control over the many elements that often accompany conversions. By trying to make conversions the domain of the SEO, we’re essentially showing our ignorance of the sales process and glorifying ourselves.

If you’re going to work for a percentage of sales and other forms of pay for performance, you’d better make sure that;

  • You have enough control (beyond SEO) to actually affect conversions
  • You have the knowledge (or partners) required to work in these other areas

Because if you don’t – I can assure you it will end badly. And really, once you’ve reached that level; why not just start your own web property? You have graduated and it’s time to move on… :0)

Anyway, I’ve stewed about writing this for more than a year, I am certain it won’t necessarily be a popular view, but it is a reality worth discussing. Also, this doesn’t mean people shouldn’t learn more about factors affecting a conversion rate, that’s also a smart move. But we should understand the domain of pure SEO and where it fits into the marketing mix…m’kay?


Until next time – play safe

David Harry
David Harry is an SEO and IR geek that runs Reliable SEO, blogs on the Fire Horse Trail and is the head geek at the... Read Full Bio
Subscribe to SEJ!
Get our weekly newsletter from SEJ's Founder Loren Baker about the latest news in the industry!
  • @steveplunkett

    i was going to argue with you a bit… then i re-read it… good article. =)

  • Barry Adams

    Throwing pay-for-performance out the window does beg the question that, with the growing unreliability of rankings as a vital SEO metric (universal search, personalised search, etc), what the key metric should be to judge an SEOs performance by. Traffic share? Overall website traffic? Keyword diversity of the incoming traffic?

  • Terry Van Horne

    Dave unfortunately that picture is what the SEO process has come to. We’ve forgotten we serve customers not Search engines. I’m of the opinion the Industry has to regress to go forward. What you have mentioned as the things that matter are what we used to call the basics. IMO, in all this time what mattered then, is still, what matters now because they are the connections and means to engage the customer/user.

  • David SEO consultant

    Yes indeed. ” If you’re going to work for a percentage of sales and other forms of pay for performance, you’d better make sure that you have enough control (beyond SEO) to actually affect conversions .” Basically the question is where SEO ends and conversions start. Wghen you alter a title tag for higher rankings, how does that affect click-thrus? And what about altering header tags and page copy? Conversions is a poor SEO metric, but SEOs should factor conversions into the equation for the ebst interests of the client, too.

  • Roger Sikes


    While SEOs should definitely watch conversions and report positive or negative trends to clients in regards to the quality of search-related traffic coming to the site, it would be extremely difficult to base a pricing on a pay for performance model, simply because as you mentioned, there are many factors beyond the SEO’s control when it comes to converting traffic.

    While it would be great for everyone for a standard pricing model to develop in regards to SEO, I don’t think pay for performance based on conversions is going to be it.

    Good post!

  • David Harry

    @Terry as ye well know I do have a pet obsession with analytics/conversions and the like. And yes, back many many moons ago I spent time learning about the many faces of marketing and how they interact. It really is a matter of defining the role. At a certain point you are venturing beyond the realm and one needs to expand their knowledge to be effective. But that’s no longer SEO. SEOs working at larger companies or in-house often don’t have the ability to affect change. As such, it is but another metric we learn from and refine tactics with. Not THE most important factor as I’ve seen touted of late.

    @ Dave – Oh most certainly. As I was just saying, it is a metric we can use. It is good for refining programs and kw/phrase targets. What we can’t do is be lulled into making it too strong a KPI. Or we fall back into the ‘rankings’ as a metric problem. And in some situations, where ability and access avail, it can be used. I just felt that those giving it too much weight or considering PFP, might not have thought it through.


  • Chase Granberry

    Developing success metrics for whatever kind of service you provide is important. As internet marketing becomes more complex people will specialize in more and more niches … eg: viral content creation, SEO copy writing, on-page SEO, link building, etc. Picking success metrics for you service is crucial to happy clients, but reporting on other metrics is important too.

    Helping the client understand how all these different metrics affect one another can really help you build those relationships, gain trust and sign more deals. Not to mention explain why you’re doing your job and someone else isn’t doing theirs 😉

  • Lyndsay Walker

    Lots of great points about how using conversions as a metric doesn’t work for SEO. But I still use it and here’s how/why.

    Traffic from organic search is not worth much unless it’s qualified or as you say, targeted. So, I use conversions as one of my metrics to determine qualified traffic (I also use bounce rate, time on site, pages per visit, etc.) But conversions is a good one since if it’s REALLY qualified traffic, they’re likely to convert.

    But the other way I use this number is as a percentage of conversions. What percentage of conversions came from organic search? The goal percentage will depend on what other campaigns the site is running, but if SEO is the main marketing effort at the time, the percentage of conversions from organic search should start to climb.

    When it comes to pay-per-performance, I couldn’t agree more. Unless you have complete control over the entire conversion process, there are too many “wild and out of control” factors.

    To sum it up, conversions is an important metric as part of the reporting process, but it should not be relied upon as the ONLY success metric.

    • Todd Heim

      I totally agree with Lyndsay. I use conversion stats to determine the quality of traffic coming from search engines. It shows me what keyword research does not.

      An example of this: a client of ours has 2 main keywords. The one phrase represents about twice as much traffic as the other, yet still only about half the conversions. It doesn’t take a rocket scientist to see that focusing on the latter is more valuable.

      While it’s not a great success metric on its own, it’s a good way to show your clients how they get a return on their investment. And paying attention to it allows you to further optimize for ROI, not just rankings or traffic.

  • Alysson

    I used to work for a large website design and marketing agency. Separating SEO from conversion optimization was a brick wall I seemed to constantly be banging my head against. Corporate powers that be wanted to roll it all into one and refused to accept that they are two separate specialties that can’t simply be combined together for the sake of “efficiency”.

    • loren

      That’s fine as long as those ‘powers that be’ then give you control of the factors controlling conversion rate and give you the resources to do so.

  • Head Alienst

    Site evaluation for conversion is an Art Form, it is based on the Psychology of the visitor, their tendencies based on historical performance and common sense. It is easily measured, and in most cases can be manipulated to a successful result based on revenue or whatever model the client requires. Now, we handle all aspects of SEO, however we have found that this dynamic allows us to properly service clients in a shorter time frame at a flat rate fee.

    For example: we have a client with 2000 visitors per day, average sale of 250.00 and less than a $1000 a day in sales. We evaluate the traffic, clean up some of the less than ideal tags, etc.. Evaluate the site structure, flow, layout and Psychology, etc. Twelve days later the site is converting with no change in traffic $1500 per day on average, by month three more than $2500 a day, Month five they broke $100,000 in sales with a 4.4% conversion ratio. (vs. .019% and $14,000 in sales the previous December) That’s all the measurement that is needed to prove the success of your program to any client, present or future.

    Its very simple, if you have traffic and no sales you need Marketing Psychology based on Conversion, etc. If you have no traffic and no sales, you need it all.

  • gregrod

    I agree they are two separate things but IMO, they are closely related to the goals/desires of almost all clients; that being an increase ROI. Therefore – SEO to improve rankings and increase traffic first, then LPO/conversion 2nd (IF you are capable of doing so).

    You’re right, if you can’t control all elements, you could be spinning your wheels for nothing. BUT I think it should be addressed where possible. Even simple things like trust labels and improved or better placed CTA’s can go a long way to improving conversion. I think most SEO’s should know the basics.

    If you have a client that you get great rankings for but their revenue has not improved do you tell them, “Hey, I’m done – you’ve got great rankings.” or do you look at analytics and find out why high traffic terms have a high bounce rate. Wrong term? Wrong landing page? Wrong message on the right landing page? Improvements to bounce rates will in turn, positively affect conversion.

  • Raquel Hirsch

    Great discussion! (Finally!)

    As a proud representative of the Conversion Optimization side of the business I can tell you that, based on over 2.5 years running conversion optimization tests for the largest variety of clients, the only way to know *for sure* if you are getting the most possible conversion is by running tests.

    In terms of what are the variables that matter, we have developed a model to evaluate the current landing pages and create hypotheses for testing. We call this model LIFT Model™ (Landing page Influence Function for Test).

    WiderFunnel’s LIFT Model evaluates the current (or Control) web page against six core performance criteria in order to create hypotheses worth testing and which are likely to generate statistically valid results in a short period of time — and criteria #3 is the linkage to Search.

    The six LIFT Model evaluation criteria are:
    1. Value Proposition
    2. Urgency
    3. Relevance
    4. Clarity
    5. Anxiety
    6. Distraction

    Relevance is all about matching the intention the web visitor had when s/he clicked on a search term result to land on the page.

    You can read more about it here: http://www.widerfunnel.com/conversion-rate-optimization/the-six-landing-page-conversion-rate-factors

  • dhiraj

    I agree on @ terry “We’ve forgotten we serve customers not Search engines”

  • Amin

    Our agency recently acquired an account which had previously been serviced by a reputable SEO company.

    Our client switched because he felt that the previous SEO firm had “misrepresented” the deliverables of the SEO engagement. In our client’s mind he was buying not only targeted traffic but also greater conversions.
    The SEO firm was focused on bringing targeted traffic, and on that metric they were successful.

    When we came into the picture, we started out by defining conversions (lead forms, inbound phone calls, ecommerce sales) and then started measuring existing keyword performance against the respective conversions. We also implemented a PPC campaign to establish converting keywords.

    Our SEO initiative is now in implementation. We are hopeful that our tactics will have paid off and 3 months from now we would have delivered to our client targeted traffic with enhanced conversions.

  • Nancy E. Wigal

    Thank you for a good case to make to my prospects. I don’t run their company, I don’t control their marketing, I don’t have control of the development of their products and services.

    I make a very clear point to tell them what I can and do control. Sometimes they still don’t get it, but that’s when I walk away. I don’t need that kind of client, especially as a solo operator.

    Great post!

  • rohn smith

    Hi ,you really observed good things.Many people including me just do some things for search engine but not for users.

  • JW

    Great article. Conversions are so critical but often frighten clients when you take them to this level and talk about including their response and customer service to the enquiry. Especially service related businesses.

  • David Harry

    Lot’s of comments, so let me try a ‘bulk’ answer. Once more, I am not against using conversions as a metric in SEO. It is obviously important in establishing which terms are performing the best and getting tighter targeting. Nut unless you are versed in conversion optimization and have the control to affect change, it isn’t the primary KPI. I wrote it as I am a fan of conversion optimization and felt not enough SEOs were qualified to be using it as a metric.

    At the end of the day, it was the spate of PFP type posts I run into and wanted to highlight just a few of the related issues with that type of approach. There are exceptions to every rule, but it is wise to consider the many elements affecting conversions beyond SEO.

    So, before you hang your hat on it as a KPI, be sure to keep this in mind.. IMHO as always…

  • Adam J. Humphreys

    We had a discussion about this at the web analytics course at UBC yesterday. There’s definitely a big difference between profitability, and conversions. Your ROI for an ad isn’t necessarily the same thing as being more profitable for the company.

    One thing I believe you should ad to this article is that professional SEO’s are often web developers prior to getting into SEO. Anyone who gets into SEO with out first understanding web development/design as well as the site life cycle has no business calling themself an SEO analyst. Site structure over all is every bit as important as the content on it.

  • John

    SEO is an “inbound” marketing function that drives qualified traffic to your website. That’s it. Customer acquisition and retention is the next step (“outbound”) where you manage the pipeline, implement drip marketing tactics and convert prospects into paying customers.

  • Thomas Bright

    Great post! Send Traffic, no responsibility or feedback on what can convert…you’re spot on! I haven’t read an article like this in a long time, way to not hit the mark. Did you optimize this article for how NOT to the SEO?

  • John Santangelo

    Of course you need to pay attention to conversion rates. If one keyword is converting poorly, then you change focus to a different one (perhaps even a slight variation of the term) and it converts well…. well um.. that’s definitely a good thing! It’s not the only metric to watch, but ignoring it will cost you or the site $$$. So what if it’s not technically SEO. It’s part of internet marketing.

    Pay for performance is another story though.

  • Alan Bleiweiss

    I get asked this more and more. And I am brutally honest right away.

    Unless I control the company I’m doing the work for all the way down to actually having control over the product and services actually being offered, let alone the marketing copy, web site design UI and content, there is no way I will ever work on a pay for performance basis.

  • Jack Adams

    Anyone without the smarts to understand what conversions as a metric does and does not explain in and of itself should be allowed to make the requisite mistakes that would force them into another industry that’s more suited to their skills (or lack thereof).

  • ashbey

    Conversions are what is clients are concerned about. An SEO can boast about getting #1 rank for #keyword but for client it would be nothing if there are no conversions.

    The “Factors affecting conversion rates” you have mentioned, most of it would be part of website analysis, pre SEO phase. An efficient SEO would be well aware of them and hence have full control of what decision he makes.

    What do you think about “Server down time”?
    Other than “Conversions” which other metrics is it going to affect?
    Almost every one, directly or indirectly.

    I do not find any reason to pick out just “Conversions” as a messy SEO metric.

  • Stoney deGeyter

    SEOs certainly can affect conversion rates and that should be a core focus. But, as you pointed out, the SEO doesn’t have enough control over it that it makes sense to make it a good payment structure. I wouldn’t stake my income on something I can’t control to that degree.

  • Raquel Hirsch

    Well, according to Randy Fishkin of SEOmoz, he present a table describing three tiers of marketing channels and their scores for Average ROI, Average Effort and Average Cost. Conversion Optimization appears in the #1 placement for Tier 1.

    So… no need to be squamish about improving the conversion rate (more comments here: http://www.widerfunnel.com/blog)

  • Jon Bentley

    Great article, and a really interesting topic that I think is going to become more and more prevalent.

    Whilst the two topics are currently separate, and as a result should be kept separate (i.e. separate skill sets, and payment for SEO should not be conversion based) I think they are going to start to overlap.

    With personalized search, rankings are going to mean less, and conversion rates (the real reason why our customers do SEO) will become more important. Therefore to prepare for the future, an increased focus on conversion optimization should be taken, and perhaps sell these addition features, becoming more of a Internet Marketing provider, rather than an SEO provider.

  • Alexander de Albuquerque

    Conversions are 100% relevant to SEO. Here is a question. What is the point of driving traffic to a site that doesn’t convert? Whos fault would that be? Did we rank for the wrong keywords or is the site simply not competitive enough?

    Conversions are a part of “Online Marketing”. SEO is simply one channel. It is directly responsible for the “quality of traffic” brough to the site. Catchy page titles, headers, content, calls to action, even site layout requires webdesigners, online marketers, copy writers, marketing and PR people to work collectively. The post above refers to the SEO servicing in a very narrow “old fashioned way”. SEO is evolving as quickly as the web does and simply “driving traffic” isn’t enough. The best SEO’s will understand the purpose of any given site and work towards achieving goals that stretch beyond rankings.

    Analysing data from Analytics to fully understand a visitors response to any given site is an SEO job, woudn’t we agree?

    SEO has evolved way beyond good rankings.

    Although I agree, you cannot possibly base an SEO campaign on a “pay on performance” basis as this isn’t feasible in an environment where SEOs have no control over rankings or visitors.

  • nightlifeblues

    yap off course Site evaluation for conversion is an Art which is based on the Psychology of the visitor, their tendencies based on historical performance and common sense. It is easily measured, and in most cases can be manipulated to a successful result based on revenue or whatever model the client requires.


  • nightlifeblues

    yap off course Site evaluation for conversion is an Art which is based on the Psychology of the visitor, their tendencies based on historical performance and common sense. It is easily measured, and in most cases can be manipulated to a successful result based on revenue or whatever model the client requires.