Conversion Optimization for Local Businesses

May 26, 2015 · Analysis

How automated bidding and lookalike data help lower CPA for small business advertisers

Most small businesses and local businesses spend money on search marketing with one simple goal: find new customers.

Of all the metrics search marketers track, cost per acquisition (CPA) is usually the most important to the advertiser. Clicks are fine, but they represent only the possibility of a customer, not an actual lead or sale. How much am I spending to make a sale, get a new customer or snag a qualified lead? That’s what most directly affects the advertiser’s bottom line.

Campaign managers and search marketing technology companies emphasize conversion optimization – doing more of what works so advertisers get more for their money. This means examining successful past conversions, looking for patterns, and adjusting campaigns to maximize the chances that the users most likely to convert will see the advertiser’s search ads.

Defining “acquisition” or “conversion”

So what, exactly, represents a successful acquisition or conversion for a small business or local business?

For e-commerce companies, a successful conversion is a sale. Assuming everything has been set up properly on the advertiser’s website (more on that in a minute), it’s relatively straightforward to connect the dots between search ads and revenue. Google Analytics and other tracking software can show exactly which online sales came via clicks on search ads, and even which keyword the user searched before seeing the ad.

This direct connection lets companies like ours create algorithms that work in the background, analyzing what happened before every sale: Which keyword brought up the ad that led to the sale? Where did the ad appear on the search results page? What time of day and day of the week did it happen? Was the user on a mobile phone, tablet or desktop computer?

The technology then looks for more users like those – assuming that showing the advertiser’s search ads to users like the ones who purchased in the past should result in more successful conversions in the future – and automatically adjusts its bids to try to “win” those users in the online ad auction at the lowest price. Reaching more of the users who are most likely to convert at the lowest possible cost per click results in a lower CPA – more successful conversions for the same total budget.

A proxy for conversion data

In theory, automated bidding algorithms should be able to optimize conversions in the same way for any advertiser. But for the kinds of small and local businesses our partners work with – plumbers, law firms, dentists, auto dealers, roofers, etc. – it’s often much harder to tie a search ad to an actual sale.

For a plumber’s search campaign, a “conversion” or “acquisition” is usually defined as a phone call, an email or filling out an online form requesting a consultation. These are closer to a new customer than just a click, but they’re still not actual customers. None of the people who call, email or fill out the form may hire the plumber, or all of them may. Most likely it’s somewhere in the middle.

Just like in the e-commerce scenario, if everything is set up properly on the plumber’s website, search marketing technology can give campaign managers keyword, ad position, day and time, and device details on the users who “converted.” This is important information because it gets the business closer to understanding the campaign’s ROI, but for the plumber, it’s still only a proxy for actual revenue.

Small businesses make educated guesses

In a perfect world, the plumber (or law firm, dentist, auto dealer, or roofer) would know exactly what percentage of phone calls, emails and form fills lead to revenue. In our experience, however, while some small businesses may be able to make an educated guess, most are not tracking these leads carefully enough to be certain what percentage turn into paying customers.

It’s not that they don’t want this level of detail – in the ideal world, of course they do. But with all the day-to-day tasks involved in running a small business, and without a sophisticated CRM or lead tracking system in place, in the real world this kind of analysis falls off the to-do list.

And it gets worse. Remember how we said “assuming everything has been set up properly on the advertiser’s website” above? Well, the fact is that many small businesses and local businesses don’t have things set up properly. They don’t know how to install tracking pixels. Their websites haven’t been updated in years. They may not even have access to their website, because a former employee’s cousin’s girlfriend set it up years ago and disappeared with the hosting information and password.

All of which has the same result: no conversion data and no easy way to get any.

Bringing ideal-world learnings to the real world

Here at MatchCraft, we know that most of the advertisers with campaigns on our AdVantage platform don’t have hard conversion data. For those who do, our algorithms use that data to boost their campaign performance and optimize conversions. But for those who don’t, we use other proxies to get as close as we can to the same goal.

First, our automated bidding and pacing algorithms look for the lowest-cost clicks within each advertiser’s target audience. If you assume that every click has the same likelihood of leading to a conversion, then paying less for clicks will lead to a lower cost per conversion, or CPA.

Of course, common sense tells us that factors like keyword, position, time of day and device probably do have some effect – but since we don’t know for sure, our algorithms start by assuming a constant conversion rate.

Second, we use aggregated conversion data from the hundreds of thousands of small business campaigns we’ve run on our AdVantage platform to make predictions about conversions by vertical. A plumber’s campaign is more likely than not to follow similar conversion patterns as other plumbers; the same goes for the personal injury law firm, cosmetic dentist, etc. In the absence of data specific to the individual advertiser, data from a similar group of advertisers is the best proxy.

Many search technology companies talk about conversion optimization, and every company approaches it differently. Our priority remains getting the best results for all our advertisers – from the most technology-savvy to the least.

RECENT POSTS