high CTR doesn't always mean better conversions, though there is often...

[GuyFromChicago]

"high CTR doesn't always mean better conversions, though there is often some correlation"

Thoughts on that comment?

While CTR and conversion rates can be related - theory being if your ad is well written (in relation to the keyword+match) and an accurate representation of what a searcher will find after clicking your ad the likelyhood of them converting will be higher - I've not seen this play out in the real world with any degree of statistical relevance.

Has anyone seen a meaningful correlation between CTR and conversion rates?

[Huligan]

Originally Posted by GuyFromChicago

Thoughts on that comment?

I can believe there is a coorelation, but I think it's so small that you can't see it unless you are running a really big account. Furthermore, you would have to be running multiple large accounts crossing different industries/niches to prove it.

Originally Posted by GuyFromChicago

Has anyone seen a meaningful correlation between CTR and conversion rates?

No. Often when I'm helping a client with their PPC, they will agree to improvements to their ads first. They won't let me touch their website until I've proven I can get people to their website (do what I said I could). So I can instantly improve their CTR, but the conversion rate stays fairly steady until I get permission to improve or create landing pages. I see this more than what you have shared.

[CustardMite]

I'd expect the opposite, at least on some search terms.

The maximum clickthrough rate would be 100%, if everybody that searched for a keyword clicked on your advert.

The more generic the keyword is, the less likely this is to be what you want. For example, if you sell second-hand cars, then you want all the clicks you can get on 'second-hand cars', but not 'buy cars'. On this latter example, a lot of the searchers don't want what you're selling.

You have to be careful when comparing adverts with different messages (rather than just different wording). Telling the world how cheap your cars are, and telling the world how good they are may have two totally different clickthrough rates, but they'll probably have totally different conversion rates as well (particularly if your cars aren't cheap, or aren't very good!).

What concerns me is the numbers in his examples.

He compares adverts that have been running over a longer period with adverts that have been running for a shorter period (in his final example). This assumes that nothing has changed over time (ad position, competitor adverts etc, seasonality, day of the week), which is highly dangerous.

He throws out an advert (again, in the final example) after just 12 clicks. 16 clicks would have been sufficient to 'save' this advert, and he's completely ruling it out? Why not wait another day or two, to be sure?

He says that the conversion rate is unaffected - how can you tell after 14 and 12 clicks? A 10% conversion rate would expect to yield 1 conversion. A 5% conversion rate would expect to yield 1 conversion.

Whilst I've previously suggested that waiting for statistical significance is over-rated (http://www.epiphanysolutions.co.uk/g...ver-rated.html), I think that you need to wait for a bit longer than he has for results, and certainly, if you've got any changes to your content (as his adverts do), you need to be very careful (http://www.epiphanysolutions.co.uk/g...nary-tale.html).

That said, I've nothing against his approach in general, it just looks rushed...

[Cobnut]

Originally Posted by Huligan

They won't let me touch their website until I've proven I can get people to their website (do what I said I could). So I can instantly improve their CTR, but the conversion rate stays fairly steady until I get permission to improve or create landing pages.

Amen to that! I always get a shudder down my spine when a client approaches me and says 'We've just had our new website finished, and now we want to advertise it'. You take a look at the site and your heart sinks...

I had one that had built an entirely Flash site with not a single page (including the home and landing pages) with a load less than 900Kb. Shudder.

Jon

[Huligan]

Originally Posted by Cobnut

I always get a shudder down my spine when a client approaches me and says 'We've just had our new website finished, and now we want to advertise it'. You take a look at the site and your heart sinks...

Been there. You know they don't want to hear they have to spend more money on it just to get a quality landing page right after finishing the build.

[pittbug]

For one of my clients I actually optimized the ads to lower the CTR. We had to raise bids to maintain ad position but it was worth it to them. In their case reworking the ad saved them a lot of time not dealing with junk leads and they can now do a better job of closing quality leads.