What Is a Local Ad in Mobile?

The folks at BIA/K have updated their mobile forecast: $2.9B by 2014 in the US. I have some critiques of their assumptions, which I won't focus on now. But there's something in the press release that raises an interesting larger philosophical question around "accounting" and forecasting in the local-mobile ad space. It's an issue I've been thinking about for the past six months and this gives me an opportunity to write about it.

BIA/K says that local will represent 69% of US mobile advertising in 2014: 

BIA/Kelsey expects U.S. mobile local advertising revenues to grow from $213 million in 2009 to $2.03 billion in 2014 (57 percent CAGR). This represents 44 percent of total U.S. mobile ad revenues in 2009, growing to 69 percent in 2014.

This is a huge percentage and it begs the question: "what's a local ad?" Accordingly this is the part I want to focus on:

BIA/Kelsey defines mobile local advertising as that which is targeted based on a user’s location and/or actionable locally. Local targeting occurs to varying degrees and with different methods within each of the advertising formats examined in the forecast (search, display, SMS).

Again: "targeted based on a user’s location and/or actionable locally." Let's unpack this a bit. 

Arguably all product advertising in mobile is "actionable locally." For example a mobile display ad for a Sharp TV becomes "actionable locally" if it prompts me to head into a retail store and look at or buy the set. It may or may not have a "local call to action." And over on Screenwerk I've argued for five years that product search needs to be considered a part of local because that's where most of the transactions ultimately occur -- in stores. 

But transaction-location swallows almost all commerical activity and some people may feel that's too broad a concept. Similarly "actionable locally" is vague. I believe what they're trying to get at however is something like a coupon that needs to be redeemed in a store or a business service that must be fulfilled offlline.

But here's an interesting hypothetical that illustrates the challenge with this idea. What about a Gap ad (discount/coupon) in a mobile app that equally applies at all Gap stores across the US? How should this ad be categorized; is it a national ad or a local ad?

It may target audiences across the US equally and it doesn't necessarily contain local ad copy (in fact it probably wouldn't at this stage). Maybe it's exclusively for in-store purchases but maybe there's an e-commerce component (which is increasingly true for retailers: channel agnosticism). There may be a secondary or subordinate link on a landing page to a store locator. Absent any other local copy does this store locator make it a local ad? (More on that later.)

There are several considerations that are relevant to defining a local ad in a mobile context: targeting methodology, ad copy and ad format. Because users who see ads on mobile devices are always somewhere that can be pinpointed quite precisely, every mobile ad has the potential to become local in a way not possible on the PC. 

Millennial Media reported that in October roughly 18% of all display campaigns it saw were geotargeted. Here "geo" is defined quite broadly to include country and state. Given that marketers can target mobile users with great precision, what level of geo is required before we call an ad local?

Picture 27

Does an ad need to be targeted down to the DMA or city level to be considered local? Or would we be willing to call ads that target France, for example, or all of New York local? I don't have an easy answer but I would argue we'd need to get down to at least the DMA level. We could call a state-level ad "geotargeted" (because it is) but "local" implies something more narrow.

Now to ad copy. Clearly an ad that contains city-level references would seem to qualify as "local."

The ad below, from a JiWire-run campaign (online), was a national buy that dynamically inserted local references to make it appear more relevant to users in specific markets. But it did this across the US; it was not otherwise a "local" ad. There was no local call to action, no store locator; it was a pure brand campaign that happened to include location references. Is this a "local" ad?

Picture 28

Now back to the "store locator" issue. Recall my "first date with iAD." I saw an ad for Klondike Bars. There was nothing local in the content of the ad, except that it did offer a store locator of sorts ("find a bar"). Is this a "local" ad?

Screen shot 2010-09-20 at 5.28.25 AM

My view is that most brand-oriented ads in mobile are going to contain dealer or store finder capabilities as a matter of course. It will essentially be a "checkbox." This is because the phone and its functionality (maps) permit it -- so why wouldn't you do it? It makes brand messages actionable locally. Buick? Find a dealer. Klondike? Find a bar. Marriott? Find a room.

If we consider these local ads then more and more mobile display moves over into the "local" column. That raises a related issue: ads with phone numbers in them.

As I just argued mobile ads (whether search or display) will routinely have store locators or links to maps. But they will also increasingly show phone numbers too -- again because of the way the inherent capabilities of the handset can be invoked. Does a national, brand-centric insurance ad buy (e.g., State Farm) become local if it contains an 800 number that routes calls to local offices? What if it has a dynamically inserted local tracking number but no other local element? 

While an ad for a local sushi restaurant is clearly a local ad (one town, one restaurant), some of these other scenarios (national --> local) are much more ambiguous. And as I suggested, location and local ad copy will increasingly be dynamically inserted based on a national database of locations, ad copy and images. Google is already doing this in mobile today. 

There's somewhat less ambiguity when it comes to search advertising but not much less.

The focus for small business will be less on buying mobile advertising per se than getting exposure broadly across platforms via channel enablers. There will be some mobile-specific activity by SMBs (e.g., Foursquare marketing, Facebook Deals) but most marketing will not be mobile-centric. Indeed, very few true SMBs will be buying PPCall ads on Google. Most of the action for SMBs in mobile will be about organic distribution.

For the foreseeable future most of the "local" advertising on mobile devices will be bought by enterprises that otherwise seek regional or national reach but local stores, dealers or outlets. Thus we return to the various scenarios above and the question of what do we consider a local ad in mobile?

It's a much harder question to answer than it seems.

Under the broadest definition of "local" the category swallows the lion's share of mobile advertising going forward. And we can manipulate the definition of "local" to make the category larger or smaller. But where we place ad revenues is less important than how consumers are interacting with mobile devices and what sort of marketing or advertising methods are effective in reaching them.