Measuring How Search Ads Drive Offline Conversions
Moderator: Misty Locke, President, Range Online Media & Chief Strategy Officer, iProspect, Range Online Media / iProspect
Nadir Hussain, COO, Media Flint, Inc.
Leigh McMillan, SVP, Marchex Voicestar, Marchex
Wister Walcott, co-founder and VP of Products, Marin Software
Vivian Yang, Senior Manager, Global Direct to Consumer Marketing, Electronic Arts, Inc.
The room is pretty sparse because Ask the Search Engines is happening in the large hall. I know there’s great coverage of that happening so I decided to give readers an alternative with the Analytics & Conversion Track.
Misty explains that Yahoo! did a study in 2007 called ROBO (research online, buy offline) that attempted to quantify “the impact of search marketing and display advertising on consumer shopping behavior and the in-store sales of major retailers.”
Misty: Why aren’t we looking at direct ROI from online efforts toward offline sales?
Wister: It’s complicated and not everyone’s in a position to observe that interaction. Direct response marketers aren’t going to want to spend when they don’t see the direct ROI. And the person that hired that person is trusting them to make the decisions because they don’t know search.
The marketer is on the one hand being held to their results, and on the other hand may know there’s something more there.
Vivian: The biggest question the ROBO study is trying to solve is, search engines are always asking for a big share of a marketer’s budget, but even though 87 percent of people are spending time online, e-commerce only is a single digit of sales. If I only generate a single-digit of sales, then I’ll only get a single digit piece of the budget pie. There’s a question of why people spend so much time online and contribute to a small portion of sales.
We need to understand the consumer’s behavior. From awareness to excitement to eventually closing the deal, it’s a long process and each touch point has a different measurement associated with it.
Misty: Where should I start?
Nadir: Start with search and add in display as well. They point to be able to track. The Internet has spoiled us into thinking every conversion and click is trackable. Because the action isn’t happening online it becomes more difficult to track. But start with search because it’s easy to track.
Leigh: Track your phone calls, including from search. A lot of phone calls for service-based businesses, and these are being driven by search campaigns, whether you know it or not. Attributing these calls to search ROI will let you spend more.
Wister is gong to share his slides, which cover different ways to look at search and offline conversions.
Online to In-Store
- Q: does on-site activity drive in-store activity?
- Experiment design:
- Create equivalent test and control geographies/products
- Drive increased traffic to test locations/products
- Compare preliminary conversion events (e.g. product detail page view)
- Compare in-store product lift
You should see that when you’re spending more ad dollars on your test, the conversion dollars will go up. Then you want to see if there are additional sales for the product inside the store.
Online to Phone
- Q: accurate value of keywords, creatives?
- Implementation requirements:
- Session-level phone number mapping
- Post-call closure
If you have a lot of keywords you have to map it to the sessions. With post-call closure look at call length (a long call could be a sign of a conversion). Also, you could have the call tech report info after the call, like whether it closed or not.
Within Online: Display to SEM
- Q: display supports brand?
- Experiment design
- Integrate SEM and ad server tracking
- Drive traffic to test and control ads / geos evenly
- Compare CTR
- For visitors that saw test ad
- For visitors that saw control ad
Nadir is presenting next. What are the different kinds of offline conversions?
- Phone calls
- Did the phone call result in a sale?
- Potential client walking into a physical store (not trackable)
- A person redeeming a coupon at a physical store
Why is it so important to track offline conversions? Offline leads are even more likely to result in a sale. Phone call leads are warmer than a form lead. It is a huge mistake to optimize your campaigns based on only partial data, i.e. based only on online conversions.
He tracks which keywords within AdWords campaigns generate phone calls. If you put it together with form analytics data, you can have an aggregated column of cost per lead, driven by keywords.
Leigh jumps in and says that she thinks of call analytics, not just tracking. Think beyond where a person came from and when. With call recording you can pull out a lot of business intelligence, understanding what, down to the keyword, is driving calls. You can improve customer service. It’s not just about tracking where a call comes from.
Audience question: What valuation are you placing on calls compared to conversions and does that vary by cost-point?
Leigh: It varies a lot by industry. In finance, a phone call and a form lead may be valued equally. With her customers, they value phone calls higher than phone leads simply because they convert higher.
Q: What’s the lag time between receiving a form and getting a call from the same customer? Also, what happens when you cancel your vendor and you want to keep your phone numbers?
Nadir: In their experience, they see that if a user submits a form and doesn’t receive a phone call from the sales team, they’ll call within 24 hours. For question two, they quarantine newly acquired phone numbers 30 days to make sure the previous life of the phone number is gone.
Leigh: All toll frees have been used at some point. Quarantine 800s longer than 877s. Plus misdials are easy with toll frees. The goals is just to minimize the noise.
Misty asks Nadir if he wants to finish his slide presentation and he says that he’s happy going with the question driven model that’s surfaced.
Q: My products, video games are driven by in-store purchases. Can you share any info on this topic as it relates to the game industry?
Vivian: Today, more than 3/4 of sales come from in-store sales, and that’s mirrored in their marketing budget. They’ve remained a packaged goods company. But if you look at the video game category, it’s changing. When the economy went soft, video games went stagnant, followed by single digit and then double digit declines. The video game segment is shrinking along with the store shelves. But they’re seeing growth online. They had to figure out what was the optimal marketing budget mix. They found a strong correlation between search interest and offline sales.
There’s a gap in different stages of a product launch, from pre-launch to launch phase to sustain or capture, and then long-tail. There’s a gap between demand and what was captured at every stage. Big missed opportunity.
The quick and easy way was to identify which metric matters. They knew that if there was one piece they were going to influence it was going to be the channel that occupies the majority of the budget – TV. So they thought, could search be tied in to TV? It’s a painful organizational exercise as different departments come together. They talked a lot about what is the marketing being used for. A TV spot is used to brand awareness. When they launched Dante’s Inferno through the Super Bowl commercial, they tracked the lift of traffic to sites and searches? Was there a higher engagement level? That was tracked down to the sales.
Dante’s Inferno has also been building a social app on Facebook. A new feature or character is released at regular intervals, every few weeks. Video game is still very much a word of mouth vertical, so getting that engagement has been a boost.
Misty: We’ve been talking about call tracking for 10 years now, so why hasn’t it taken off?
Leigh: People are doing it more, though I think it was slow to take off because of an over-focus on tracking online conversions right around that same time. There also wasn’t a scalable way to move someone from an online search to a phone call. With Skype and the iPhone, calls and search traffic is converging. Price for call tracking has come down in the last year, even, and the cost may have been prohibitive in the past.
Leigh’s slides are going up now.
Wister: You may want to try “poor man’s call tracking” if that’s you option. Send an e-mail after the call. The e-mail is pretty sparse, but with a URL. When they click through you can capture a cookie and see what keyword they converted on. (I don’t know that I caught all that right since it doesn’t make sense… Chime in if you know what that’s supposed to say! *Nudge, nudge, Wister* )
Q: What are the top sources for staying up on search to offline?
Nadir: Watch what Google is saying. With click to call ads, Google will be tracking that themselves.
Leigh: Matt Booth, an analyst previously at Citysearch, knows a lot about calls.
7 Replies to “Measuring How Search Ads Drive Offline Conversions”
One major issue we have is building on the brand of our clients either the name is not great or the product they push is to vague. Thanks for the read up.
Can you elaborate a little on the “poor man’s” call tracking option? How would an email track the call source?
“Poor Man’s Tracking” – I like that. It’s actually a relatively easy concept, and simple to implement. Here’s an example:
1. Customer clicks on a Google paid search ad, comes to your site, and chooses to call your sales department and converts over the phone.
2. At this point, Google has cookied the user with the details of their search, and how it relates to your account. Because the user never hit your conversion page, the conversion is never recorded in your AdWords account.
3. The solution is to send a confirmation email to the customer after they convert, that encourages them to click a link and visit a specific page. “Click here to check the status of your order.”
4. The page that this link goes to should also contain your “conversion tracking” so that when the person above visits, their existing cookie is activated and tied back to your AdWords account.
Of course you can add in other tracking as well for Yahoo, Bing, etc…
Feel free to reach out to me if you have more questions.
Wow that’s actually rather ingenious.
It seems three things need to happen for it to work though:
1. Your sales people need to remember to send out the email after each call.
2. The customer has to be motivated enough to click it, but the page has to be generic enough to be used for everyone so you can set up the tracking script once.
3. The customer needs to have his cookies intact.
However if those things are all in place it seems like it would be pretty effective.
Here are some helpful tips Avi…
1. We automate ours. We actually modified a pre-existing confirmation email. Sales people are great but it leaves room for manual error if you rely on them.
2. Exactly. We use the “check the status of your order” message along with a “bonus offer” – it works well. That page has generic messaging. Email me if you would like to see a sample.
Using this method has allowed me to properly source about 45% of previously unidentifiable sources.