Error message

User warning: The following module is missing from the file system: domainPrefix. For information about how to fix this, see the documentation page. in _drupal_trigger_error_with_delayed_logging() (line 1128 of /home/sites/jellyfish-drupal/public_html/includes/bootstrap.inc).

SEO Forecasting is Guesswork at Best

Insights | 16 Jul, 2013

In this Jellyfish Insights video, Matt Owen, Head of UK Agency, talks about the challenges of SEO forecasting.

Video transcript

Hi. My name is Matt Owen, and I'm Head of UK Agency here at Jellyfish.ξToday, I'm going to talk just for a few moments around SEO forecasting andξwhy it's maybe a bit more challenging than you might imagine.

So, obviously, as a business we would love to provide you guys with an SEOξforecast. It helps you to plan. It provides clarity and insight into what'sξgoing to go on. But unfortunately, there are some challenges aroundξassembling SEO forecasts, and that's what I'm going to run through in thisξvideo today.

And I'm going to focus on just three main data points. So the first one isξtrying to get a handle on what the search opportunity is. The second one isξtrying to understand what kind of visibility we can drive for a particularξkeyword or a set of keywords. And the third one is how much traffic willξthat actually drive to your website, in other words, what kind of clickξthrough rate could you possibly estimate?

So the first question is around search opportunity. So as you know, ifξwe're thinking about an SEO campaign, we're generally trying to look at aξbundle of keywords, and we're trying to find out how much potential searchξvolume those keywords can drive for you. So how do we do that? Well, likeξmost people, we use the Google AdWords Keyword Tool. So the obvious firstξquestion is: Well, how accurate is that tool? How well does it estimateξthe search volume that's available to you? So we've done some maths.

So if you use Google AdWords Keyword Tool, and we're looking at a specificξkeyword here, which is PPC agency, a very important keyword to us,ξand youξlook at the local monthly search volume for that keyword on exact match forξGoogle UK, you will see that it gives you an estimated traffic volume ofξ1,000 searches per month. So that sounds quite good.

So what we've done is we've had a lock in our PPC campaign, and using theξimpressions that we see against that keyword and the impressions shareξanalysis, we can actually derive a much more accurate figure for the realξsearch volume for that keyword. And guess what? There's actually quite aξbig discrepancy. So if we analyse AdWords and impressions share, we canξactually see that the monthly search volume is about 667 searches a month.ξIn other words, the AdWords Keyword Tool is out by about 49% or so. So justξhold that thought, because imagine that kind of discrepancy magnifiedξacross a whole bunch of keywords. That's quite a big data error rightξthere.

The second thing to bear in mind is when we're looking at the SEO forecast,ξwe need to try and estimate what kind of visibility gains we can deliverξacross your keyword set. Now there are some data points we can use forξthat. So what we can see here is an analysis from the SEOmoz tool, theξSEOmoz Keyword Difficulty Tool, around the keyword "PPC agency." So theξgood news for Jellyfish is that we're the number one slot here in theξnatural search ranking, and looking at these metrics, you can see that theξgreen blocks, which are around page authority, and the red blocks aroundξdomain authority indicate that we should be in that number one position.ξThey're basically looking at our backlink profile.

But what's slightly more perplexing around this is you could look at thatξnumber two slot, the agency at that number two position have completelyξdifferent metrics to we have, much, much weaker, but rather paradoxicallyξbetter than the metrics for the agency who are in the third slot.

So that's very interesting. It means that when we try to look at that SERPsξpage and guess how much activity is required to actually drive us into aξcertain position, those metrics are not very accurate. And it means thatξguessing where we can get you to is, well, it's a guess basically. Soξthat's another challenge.

So add that together. So far, the search volume estimation is quiteξdifficult, and actually estimating where we can get you to in terms of theξSERPs page is also quite challenging.

The third point that we're trying to look at here is essentially: At aξcertain position, how much traffic will we drive? In other words, what isξthe estimated click-through rates at a certain point on that natural SERPsξpage?

So we can see here a graph, and this is an amalgamation of various click-ξthrough rate estimates that have been put together byChitika. This tells usξthat if we're at position 1, our click-through rate should be around aboutξ30%, maybe a little bit more. So again, we've done some real data analysisξon this, and if we look here, this is some data from Google Webmasterξtools, and this is showing us a click through rate for the keyword "PPCξagency," which we're at position one. And just look at this again. That'sξactually says 2%, not 30%. So that's a really big variation in terms ofξthat click-through rate estimation.

Now that's for a non-brand term. So let's look at a brand term, and this isξfor the keyword "Jellyfish." We're at number two on Google UK for thisξkeyword. And again, this click-through rate is much, much lower thanξanything that's anticipated by the graph that we saw a few slides earlierξon.

So let's just put that together. So in other words, we have a big varianceξin terms of search volume demand, a very difficult job in terms ofξestimating where we can get you to on the SERPs landscape, and then, toξcompound that, a really massive unknown in terms of the click-through rateξaround that. So trying to estimate the traffic that's going to come throughξto your website is really, really difficult, and this is just for oneξkeyword. So imagine the potential data variation across a basket of 50 orξ100 or maybe 200 keywords. It's extremely difficult.

There's one more factor that I'd like to draw your attention to, because weξcould, of course, look at the data of keywords coming to our current site,ξor traffic from those keywords, using Google Analytics. But again we have aξchallenge here, because as you know, much of the specifics around theξkeywords that drive traffic is now hidden from us, due to the "notξprovided" label in Google Analytics. Here's an example of this. Andξactually for our site, we actually don't know which keywords are actuallyξdriving the most traffic.

So it is a conundrum. So SEO forecasting is extremely challenging. Soξbasically, this is a walk-through of why. I hope that's been interesting.

Thank you very much.

Please enter a valid Name.
Please enter a valid email address.
Please enter a valid Phone.
Please enter a valid Company Name.
  • Select Service
  • Analytics
  • Brand
  • Consultancy
  • Conversion Rate Optimization
  • Display
  • DoubleClick Partnership
  • Email
  • PPC
  • SEO
  • Social
  • UX
  • Video
  • Websites
  • All Services