Conversion Rates and the Marmite Effect

Did you read that post title correctly? Marmite? Yes that’s right. Conversion rates seriously have a marmite effect about them. You can end up loving or hating conversion rate metrics, and it really depends on how you apply them to a real situation. One of the biggest issues I find with new CRO clients, is that they’re not always applying appropriate metrics to their decisions. In today’s post, I’ll be discussing how conversion rates can be an absolutely horrible metric leading to sub optimal decisions, and how to turn that situation around to make the right calls.

Used inappropriately, “conversion rate” can be a horrible metric.

Conversion rates can be really misleading when used in isolation, aggregation or without a proper understanding of their context and ecosystem. It’s all about actionable outcomes, and a “conversion rate” alone doesn’t always tell us a great deal. So, what do we mean by “ecosystem”, “isolation”, “aggregation” and “understanding”?

1. Isolation

Using conversion rates in isolation is useless, it won’t tell you anything meaningful apart from the fact your conversion rate is at X% a certain point in time. Having a metric such as revenue next to it is much more insightful. Think of it this way, if ‘A’ has a high conversion rate of 10% and ‘B’ has a slightly lower conversion rate at 9%. Would we do more of ‘A’? The short answer is no, we haven’t taken into account any other vairables, such as cost. For example ‘A’ could be relying on costly promotions and offers which cannot be maintained in the long run.

2. Aggregation

Using conversion rates in aggregate is useless. “Overall” conversion is not an insightful or meaningful metric. If you segment by different sources for example SEO, PPC, mobile etc. each source will have a different conversion rate. So failing to appreciate this and relying on aggregate conversion rates leads to sub optimal decisions. Using an aggregate / site conversion rate to tell you what’s wrong with your site is like waiting for your cat to bark.

3. Understanding – higher conversion rate does not always mean higher performance.

Let’s imagine that our boss asks us for a progress report on two marketing campaigns we deployed at the same time period. We deployed strategy ‘A’ and strategy ‘B’ which delivered different conversion rates.

Strategy A = 10% conversion rate

Strategy B = 25% conversion rate

When strategy ‘A’ was deployed we achieved a 10% conversion rate and when strategy ‘B’ was deployed we achieved a 25% conversion rate. If conversion rates were the ultimate metric then basic common sense would dictate strategy ‘B’ is better since it has a higher conversion rate. Should we go with strategy ‘B’? No, take a look at this example why…

Strategy A: 5000 unique visitors and 500 conversions = 10% conversion rate

Strategy B: 1000 unique visitors and 250 conversions = 25% conversion rate

On analysis of the conversion rate formula, strategy ‘A’ has double the conversions and is better for the company and had we gone with strategy ‘B’ we might be making a massive mistake. Therefore a higher conversion rate does not necessarily mean higher performance. This is a simple example, but it really does happen!

4. Ecosystem

Conversion rates can be unreliable if the traffic to your site changes, because of seasonality, content popularity or a news event. Let’s go with the example that your content goes hot – you’ve written a series of blog posts that are absolutely awesome and reveal the secret ingredient behind the colonel’s KFC chicken, secrets behind David Blaine’s Magic tricks and the true meaning of life.

Your blog posts reel in the visitors by the thousands. That’s great! You’re more famous, your content is obviously more engaging, but your conversion rate will go down assuming the number of goals remains constant. To improve your conversion rate you would have acquire more leads but in actual fact those visitors may have only come to read your blog and may not want the product you offer right now.

Used correctly, the true value of conversion rates are exposed.

A conversion rate’s true value is exposed when we utilise it with other metrics and segment it by different sources, visitor type etc. By doing this we get a more meaningful metric that leads to an actionable outcome.

1. Indicator of performance

Conversion rates provide an indication of how well you are capturing your visitors. It is important to know how many people are converting. Most websites conversion goals tend to be sign ups, sales, downloads etc. By tracking these we can monitor the ratio of people that did convert as opposed to those that didn’t.

It can provide you with areas of concern i.e. it can tell you where your marketing efforts are most successful when segmented. If you measure it along with other business metrics such as revenue, profit per visitor, cost per conversion etc. you gain deeper insights.

2. Spur action

To some extent, low conversion rates can even spur action by senior execs. For example it can lead to improvements to the persuasive architecture of the site. The content on the site is improved which is then tested in a split test. Conversion rates can therefore spur change for the better.

3. Utilising other metrics

Conversion rates do have to be segmented in order to expose its true value. Metrics such as abandonment rates and completion rates are much more insightful in determining areas for improvement.

Let’s look at this example…

completion rate

Imagine we run a pet nutrition company and we advise people on how our products improve their pet’s nutritional intake. We’re really interested in the number of sign-ups to our site. Task completion shows you why people came to the site and how easy it was for them to complete a task but also what was achievable in terms of conversion rate.

The chart shows that 35% of people sign up to our service. So if our conversion rate is at 1% then what happened to the other 99%? Is that 99% lost sign up conversions? No, in fact it’s 1% out of 100% out of the 35%. The 35% shows you what’s achievable in terms of sign up conversions.

We also have to ask ourselves, out of all these other segments what’s more likely to convert into sign ups? So for example, of those people that come to do research, can we entice them to sign up to our services? This type of analysis is fundamental to the success of improving site conversion and understanding what’s achievable, as people applying for a job on the site might not want to sign up to your service and you may be wasting your efforts.

Think “how should I be using conversion rate effectively?”

Conversion rates are an excellent metric when used properly. Using it in isolation and in aggregation can lead to sub optimal decisions. Combining conversion rates with metrics such as revenue, completion rates, outcomes and unique visits reveal a deeper pool of insight which, in the long run can be more beneficial when you’re planning to truly make a difference to your bottom line.



Stay Updated: Sign Up for Webinar & New Blog Alerts

9 thoughts on “Conversion Rates and the Marmite Effect

  1. We like to keep a segmentation dashboard for different parts of a site and for the major traffic referrers. Conversion rate for each segment and traffic source is crucial.
    And you’re right, diving even deeper into the data before making decisions can save you loads of frustration, time and money.
    Great insight Fabian, thanks!

  2. Insightful post Fabian – I think a lot of SEO’s are able to gloss over conversion metrics, as they’re most commonly targeted on traffic and rankings. Of course this is changing as CRO becomes more accepted. I’ve been really surprised, a few times recently, on just how disconnected “conversion” metrics have been to actual data in real client reporting situations. This is a thoughtful and timely post – good work!

  3. Dr. Pete says:

    I see (I)(3) on the PPC side all the time. Some 3rd-party provider will be boasting of massive conversion rates, only to find out that they’re only bidding on brand terms. Great for conversion sure, but they’re leaving a ton of leads on the table.

  4. Hi Pete – awesome to have you here! I was surprised at the response to this post – it’s been sort of a “so it’s not just us!” penny drop moment :-)

  5. Dr. Pete says:

    Oh, I read SEO Gadget all the time – I’m probably just guilty of not being the active commenter I once was. Just so much to digest in so little time these days.

  6. I’m with you Pete. I’ve recently decided to skim out some blogs from my RSS reader, so I can have more time to participate on a select few. SEOgadget is one I can’t stop reading :-)

  7. Aw, *warm glow* CC @scientificroi

  8. John Trimble says:

    I could agree with you more, ive been in some agencies where often my bosses would get hung up on conversion rate not taking into account that conversion rates drop when the number of visits increase. Essentially Reading for many a analysis or SEO.

  9. Michael says:

    Excellent post and plenty of food for thought and testing.

    As you said taking the CRO in isolation often gives the wrong picture. E.g. Client sees a slightly lower conversion rate than last year, but forgets that their cheapest product has doubled, or that the length of the contact form has been doubled to weed out time wasters, or in a PPC campaign if you run two ads, one gets a slightly better CR, but the one with the lower CR gets a much high CTR and generates a better revenue spend. Head melter.

Comments are closed.