Why We Need Control in PPC

Last week, Google announced that the option to exclude close variant match types was going away in September. The PPC community, predictably, went ballistic. But then some began to question why we are all going so crazy – after all, how many of us actually exclude close variants anyway?

That’s beside the point, though. The point is that we want, and need, the choice.

Google has gradually taken choice away from us: In 2012, we temporarily lost the ability to rotate ads evenly, although that came back after an outcry from the PPC community. In 2013, with enhanced campaigns, we lost the ability to have separate campaigns by device, something that was once considered a best practice.

Now Google is taking away close variant options. And there is good reason to be concerned.

Brad Geddes of Certified Knowledge wrote a good article on why forced close variants is a terrible idea. He shows, with actual data, that close variants do not perform at all the same as traditional matches; in fact, in his examples, close variations have much lower conversion rates and much higher CPAs than their actual match type.

Sam Owen at PPC Hero also showed actual data on close variants, and his data was similar to Geddes’, although the differences varied by account. For some accounts, the difference was minimal, and the residual conversions gained from using close variants were worth the slightly higher CPA.

But for some accounts, especially lead generation accounts, the difference was significant, with close variants resulting in overall CPA increasing from $91.01 to $111.76. As Owen points out, this difference could be enough for a client to get upset and even decide to stop advertising with Google.

Mike Roberts at the SpyFu blog points out another huge concern about losing the ability to exclude close variants: the fact that close variants essentially turn phrase match keywords into broad match keywords.

SpyFu’s data shows that “If you spend $100K per month on AdWords, and all of your keywords are Exact Match, then you can expect to spend about $102,400 after ‘Close Variants’ is switched on at the end of September. If your account is organized by Phrase Match, then expect to be spending $117,200!” Wow.

I didn’t just take these esteemed authors’ word for it – I ran my own analyses on some of our clients. I was able to validate their findings.

The first analysis I ran was for a very large lead generation client. For exact match close variants, we saw an 11 percent higher CPA than on the exact match search queries. For phrase match close variants, we saw a whopping 35 percent higher CPA.


We also saw a higher CPC for the close variants than for the matched keywords.

All that said, for this client we are getting 20 percent more conversions by using variants, at only a 5 percent increase in CPA. So, on the surface close variants seem worth it.

However, several of this client’s campaigns are budget-limited. It seemed to me that if we excluded close variants, we would get more conversions without sacrificing CPA. So I ran the numbers.

 Scenario Impressions  Clicks  Cost  Conversions CPC  CPA  Conv % 
Without close variants  201,583 11,937 $64,171.41  142 $5.38 $451.91 1.19%
With close variants  290,112 14,630 $80,441.72  170 $5.50 $473.19 1.16%
If close variants excluded  290,112 14,630 $78,648.55  174 $5.38 $451.91 1.19%

If we excluded close variants, we could get more conversions at a lower total cost by putting the entire budget toward traditionally matched terms. For a budget-limited advertiser, this is significant. Who wouldn’t want to get more conversions for the same cost?

The bottom line here is, we need control. As Sam Owen’s data showed, for some advertisers, close variants result in more conversions for little to no additional cost. In e-commerce, for example, casting a wider net is valuable, since it’s nearly impossible to bid on exact matches for every product you sell.

But for other advertisers, close variants are costing them money, especially on phrase match close variants.

Why not give us the control we need?

To further illustrate why we need more, not less, control, let’s take a glance at the impact ofcombining tablets with desktop for the lead generation client I profiled earlier.

Device  Impressions   Clicks Cost  Conversions CPC  CPA  Conv % 
Computers 5,225,330 39,808 $122,578.89 206 $3.08 $595.04 0.52%
Mobile devices with full browsers 213,322 1,255 $7,675.78 97 $6.12 $79.13 7.73%
Tablets with full browsers 1,178,831 11,312 $16,355.79 17 $1.45 $962.11 0.15%

This client is using mobile click to call, and they get a lot of calls. That’s why the mobile conversion rate is so good. But look at tablets. The CPA for tablets is a whopping 62 percent higher than the CPA for computers. Don’t you think I’d rather use that $16,000 to buy more desktop clicks?

An easy solution here would be to add a tablet modifier. We could still bid on tablets, but set a bid that makes sense.

And how about the data for search partners vs. Google.com?

Network  Impressions  Clicks  Cost Conversions  CPC  CPA  Conv% 
Google search 1,064,870 16,231 $97,734.20 211 $6.02 $463.20 1.30%
Search partners 1,045,007 4,015 $21,250.66 43 $5.29 $494.20 1.07%

Well, this isn’t terrible. The CPC for search partners is lower, but so is the conversion rate. CPA is only 7 percent higher for search partners. But the total spend is pretty high.

A bid modifier for search partners would be great here. We could bid just 7 percent less for this traffic and be in good shape.

In fact, there are times I’d be willing to pay more for search partners! Here is data from a different B2B client:

Network Impressions Clicks Cost Conversions CPC CPA Conv%
Google search 32,606 1,057 $2,024.69 132 $1.92 $15.34 12.49%
Search partners 23,458 262 $333.76 44 $1.27 $7.59 16.79%

Search partner conversions at half the cost of Google search? Give me more, please! In this case, I’d be willing to bid 50 percent higher for search partners. Google is leaving money on the table in this instance.


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s