I just got a question in email about what percentage of searches use a geomodifier. As I was writing a response it occurred to me that a “game changer” in this whole tiresome discussion — how much of search is local or has a local intent? — is the following slightly different formulation:
What percent of search results in a local transaction?
This would shift the entire discussion to what really matters — the place where the money is spent: offline. I have been trying to argue for over three years that local is really about consumer behavior and the point of sale. Lots of people still neglect to see this connection, between a search online and an offline transaction.
If we take the question or formulation above — What percent of search results in a local transaction? — seriously then the answer goes from the 12% or 13% number that comScore uses to 90% or 95%. Looking at geomodifiers or even the more expansive “local intent” concept largely misses the big picture.
Those 90%+ figures start to change the nature of the discussion and start to very concretely show advertisers, especially larger marketers and national brands, that they need to focus on or at least complement their online advertising efforts with messages that lead people to local stores/shops/retailers/dealers in particular markets.
In my view (obviously) this is the right way to frame the discussion going forward: What percent of search results in a local transaction?
Thoughts?
Update: Should qualify all of the above in terms of commercial queries.
May 7, 2009 at 3:18 pm |
Well, it seems if you build on that baseline from comScore with what Google is doing to translate some searches without explicit geomodifiers (restaurants, doctors, lawyers) I can easily see that figure jumping, maybe doubling (not quite). I think you could look at this on a segment-by-segment basis for a deeper analysis. For example, I can see how searches for movies, books, and television shows (media searches) can arguably be traced to local transactions. But it just seems like quite a stretch to go to 90%. Looking at my own recent searches for Viking Ships and how to say “I speak French like a Spanish Cow” don’t quite connect to a local transaction – although I would be very interested in discovering how.
May 7, 2009 at 3:32 pm |
Michael: Almost all services are fulfilled offline. 90%+ of product purchases are offline. The internet plays a larger and larger role in both categories but e-commerce is flat and less than 4% of US retail. See: http://www.comscore.com/press/release.asp?press=2808
If you focus on commercially oriented queries and you look at transactions, the numbers are at that level.
May 7, 2009 at 4:21 pm |
I agree that you’ve framed the real question in the right way.
To clarify, by “local transaction” I take it you mean a transaction at a local “real world” store, as my 12 year old would say?
May 7, 2009 at 4:23 pm |
Yes, offline
May 7, 2009 at 4:28 pm |
@Michael – “Well, it seems if you build on that baseline from comScore with what Google is doing to translate some searches without explicit geomodifiers (restaurants, doctors, lawyers) I can easily see that figure jumping, maybe doubling (not quite).”
Mike Blumenthal actually speculates that it might be closer to 60x for those kinds of searches
It would be interesting to know what % of overall search is taken up by those kinds of generic service-based queries. That would give a more accurate percentage of local intent, IMHO.
@Greg, I think your ROBO-oriented figure is just about right & it would be nice to see larger brands start to recognize it.
May 7, 2009 at 4:29 pm |
I think if we collectively start focusing on and emphasizing transactions we’ll start to open people’s eyes
May 7, 2009 at 4:37 pm |
Greg — based on your 90%-95% assertion, it seems like you’re really answering the question “What percent of LOCAL search queries result in a local transaction?”.
Otherwise, you’re implying that all searches result in a transaction (95% local, 5% online). Maybe I missed something…
That point aside, I agree with your framing, and your suspicion that the number is high. Because to be fair, we shouldn’t just count the single, final search I did that returned the phone number of the store I bought my new refrigerator from; we should count all of the 57 searches I did in researching models and comparing local retailers as “searches that resulted in a local transaction”.
May 7, 2009 at 4:42 pm |
Greg: I have involvement in an industry that traditionally has claimed a roughly 1 out of 4 conversion ratio. That was pre web…and remains consistent post web. It is intrinsically local and in its fullest scope a significant industry nationwide.
I don’t believe the industry conversion ratios have changed significantly if at all post web. The simple fact is that the contacts are dramatically via the internet (in all its forms) versus preweb days (print/primarily newspapers) and other local media.
It doesn’t significantly matter whether the leads come in via MC, search, links, CL, an IYP, etc.
As in the older days, some sources do convert better. YP books often had terrific conversion rates. Its not because they were YP’s….its because they provided a platform advertising milieu that facilitated the directness of the inquiry; ie I need to find a local landscaper (and not having referrals or word of mouth experience…where else am I going to find one.). The alternative might be to see a local landscaper on local tv ads….but while it may make you aware of that landscaper….if you don’t need one then…who cares.
Search ends up being the rough equivalent of the old YP’s. Of course search is broader. One could look up landscaper in search to do research, learn about the profession….or one could be looking for an orthodontist in your neck of the woods.
I suppose a question might be….have internet buying habits changed the nature and percentage of inquiries to actual purchases in a local environment?
BTW….do you have any recent stats on % or actual #’s of inquiries that include geo modifiers…and do you have any updates on inquiries that would tend to be local…but don’t have geo modifiers.
Here is an interesting and I know rare anecdotal piece of info. I have a local smb that has ranked very highly in all 3 engines for years for the industry term w/out geo modifiers. Luck/starting early, excellent url, tons of links….anyways it ranks highly.
For long periods of time it ranked #1 for the industry term in both Y and for somewhat less time it ranked #1 in MSN/Live. Tons of traffic….but not a great converter. Its a local business afterall and in an east cost metro region. Searchers from Utah, Cali, etc. would never convert.
It has ranged between 2-8 in G for years. From Sept to Dec last year it ranged from 2-4 in Google for the industry term…and traffic for the term skyrocketed. It was a little better converter then.
As of late Jan…Google inserted a local map for the industry term (no geo modifier) at around the 4th position in the rankings.
Boom. G traffic soared and conversions for the industry term w/out geo modifier soared. High visibility and tremendously relevant to the local region.
Over the last few days it has ranged between 2-4 in data centers for Google for the industry term. That means, high organic rankings, #1 placement in the version of a localalized map for the industry term…and btw…consistent #1 position in the industry term (no geo modifier) in Adwords.
The big difference is that the organic ranking…with the industry name in the title…and the location name in the title sat way above the fold and above the map insert.
Boom, boom, boom. Traffic really soared for the industry term (no geo modifier) as did conversions. In fact….using adwords tools that identify impressions for this [exact term] search for the fairly region wide ppc/adwords campaign…the site dominated the % of traffic by far in the region. Now the region is wide enuff….that the versions of local maps (at position #4) could easily show competitors that were clearly closer to certain users…North, South, East and West. Regardless it has been a terrific boon.
I wouldn’t strive on other local businesses to try and get that self same uber high ranking for the industry term. Its probably time and cost prohibitive today for a local business…..but damn….it is a traffic and conversion bonanza.
Even w/ that traffic bonanza….the aggregate of local oriented searches (with geo modifiers) out numbers that volume of searches. Still it is a very high number for the time being.
Would still appreciate any updates on #’s of searches or %’s of searches if you can access that info.
thanks,
Dave
May 7, 2009 at 4:43 pm |
Mike: my question is rhetorical to a degree, yes.
May 7, 2009 at 5:12 pm |
Or look at it this way.
What percentage of consumers go online to search before they walk into a local store?
We build a lot of sites for local retailers and in almost any durable goods industry 70% of all consumers go online to do research before they come into the local store. It’s a consistent percentage for washing machines, sofas, lawn mowers and more.
May 7, 2009 at 5:15 pm |
Yes. Another interesting angle.
May 7, 2009 at 5:17 pm |
Good angle. Much more actionable for businesses. Seems the determination of local intent is important if you are creating search algorithms, but not very actionable for a retailer. I do think it needs some modification to “what percent of ‘commercial’ searches result in a local transaction”?
If you figure ~65% of total searches are non-commercial, and apply to that a high percentage like 90% as local transactions, then you get about 55 or 60% of all searches leading to a local transaction. Still a gi-normous number. There are a lot of queries such as, “poems to read in a wedding” where the searcher is looking for ideas and inspiration. While the wedding is certainly local and will entail a lot of local transactions, I’d be hesitant to assign that query to a transaction. I think that search is fairly non-commercial. The person is most likely to click on a non-paid organic link taking them to a page of wedding poems and not buy anything. Mr. Bauer’s “How to speak French like a Spanish Cow” query falls into the non-commercial bucket as well.
Probably should note that e-commerce accounted for only 3.3% of total U.S. sales (Source U.S. Census) which supports the high local transaction figure. Of course, a lot of that is food, gas, and other things that probably generate few searches so maybe the local transaction figure is a little lower than 90%. But still a large number.
The implicatioins are huge of couse as you note. Search has replaced much of the yellow pages and newspapers role in providing consumers with information to purchase locally and every merchant needs to be search savy. Really savy.
Might be enlightening and tangible for merchants if you took the monthly U.S. query figures, applied the local transaction percentage to it and then divide that into populations like DMAs, counties, merchants, or even homes. I just came up with a really big number for my DMA. Impossible- for-a-media-planner-to-ignore-big.
Aside: this must be a really futurist topic as all the posts are time-stamped for this afternoon.
May 7, 2009 at 6:14 pm |
May 7, 2009 at 7:06 pm |
I love the angle Greg. It’s all about re-framing the value of the web to local businesses so that they feel comfortable taking the plunge and spending real money advertising and engaging online.
In some very basic research that we’ve done on the bar vertical for example, an owner is willing to spend $5000 a month advertising in Village Voice but was not willing to spend a dime online, that was until someone walked in with our iPhone application and “proved’ to them that this medium translates to foot traffic.
What are your thoughts on how to broadcast this message in mainstream media and ultimately to the local business owners? Would be aweomse if Brad Stone did a follow up piece on hyper-local and talked about this
May 7, 2009 at 10:06 pm |
Re Brad Stone . . . that’s up to him obviously. There needs to be consistent messages from a range of sources on this subject. I’m doing my part of course. And I’m on tour
May 8, 2009 at 11:53 am |
[...] was an interesting discussion yesterday on Greg Sterling’s Blog about the number of searches that have a local modifier to them and then Greg’s twist to try [...]
May 8, 2009 at 2:41 pm |
Greg:
I want to try and address your comments more succinctly. We have businesses that address a sector of a large service industry. The industry is primarily local. The industry has been around for decades, and we have operated some smb’s within a sector of the industry for a couple of decades.
Before heavy usage of search for local business services…this industry generally reported that 1 in 4 requests tended to end up in a sale.
The industry is currently heavily dependant on the web. Its main focus of advertising/marketing awareness is based on the web.
The industry still sees roughly 1 in 4 requests turning into a sale.
Now within a certain context…requests for information that come from certain media other than search convert at a ratio less than 1 in 4. On that basis…yep…possibly search is a better dynamic for creating transactions, in that with print converting at less than 1 in 4….search must be converting at a ratio that is somewhat higher.
May 8, 2009 at 2:43 pm |
Dave:
Theoretically the internet is cheaper than traditional media. Has that been your experience?
But otherwise users are online so SMBs have to be there and address this market.
May 8, 2009 at 5:40 pm |
Greg:
Absolutely and by a very significant difference the web is cheaper than traditional media. You get up a web site. Better work real hard at getting it ranked for the right phrases. Work real hard at getting it ranked in maps and Y local. If the rankings are strong….on going costs are the price of hosting. pfft. that is relatively nothing.
We spend a pretty fair sum on ppc. PPC costs don’t come close to the costs of our experience in traditional media. In the past our primary reliance, after years of testing, were a couple of print sources….and very wide spread use of YP in our metro region. Ugh…Yp books. so many ways for them to hit you with charges for different books…and different presentations. UGH (not that it didn’t work…it was just crummy method for billing in a large metro.
I view the best search terms…like a mega old YP. The consumer has intent…searches on the web….and you better be there. Its far easier than using the old YP….so it does generate more traffic.
2ndary phrases, in search, or those that are not as focused on the immediate purchase intent…are similar to other forms of media. The user might see the info…but doesn’t necessarily but it at the moment. It makes an impression though.
The endless and various sources of advertising on the web through various outlets such as MC, the IYP’s etc. etc. Personnally I think there value is limited subject to whether you outrank them in search in the major engines. Our simple experience was that when those other sources outranked our sites in search, we paid to get traffic from them. Once we started outranking them for the most appropriate terms….they didn’t deliver traffic anymore.
For various industries where verticals dominate organic rankings, such as hotels, restaurants, etc…..these vehicles are critical, unless and until a business/site can outrank them.
Of course map inserts have changed that dramatically.
Finally, of course there is ppc. Its eminently valuable.
I’m still not convinced that local search has generated more sales…than preweb. I do believe though it is infinitely easier to research online and then purchase. Possibly in that context it has changed the dynamics of local business activity. I’d like to see research on that component of the issue.
May 15, 2009 at 4:01 pm |
[...] was an interesting discussion yesterday on Greg Sterling’s Blog about the number of searches that have a local modifier to them and then Greg’s twist to try [...]
February 10, 2011 at 1:30 am |
[...] was an interesting discussion yesterday on Greg Sterling’s Blog about the number of searches that have a local modifier to them and then Greg’s twist to try [...]
February 1, 2012 at 5:09 am |
WordPress backlinks do follow…
[...]Game Changer: ‘Local Transaction’ Percentage « Screenwerk[...]…