One of the reasons why search is so great is its accountability. The same cannot be said for some forms of offline advertising. The good news is that you can use search to help gauge offline success.
Search marketing and the metrics associated with it are able to give you a “finger in the wind” suggestion as to how well an offline marketing campaign is performing. In most cases, that is an improvement on other measurement techniques.
A simple measure of the impact of an offline campaign is the traffic driven by your brand keywords. If a billboard, radio or TV advert has encouraged people to think about your brand then you should see more people searching for your brand. Of course, this assumes you can be found for your brand otherwise these searches will not translate into traffic.
Many web analytics packages are able to show traffic in hourly intervals. If you know when your TV ad is due to air then you will be able to measure the impact on that hour and the one after.
Brand Hybrid Traffic
Hybrid traffic combines a brand term with either generic or specific terms. For example, if E-consultancy runs a TV campaign designed to increase awareness of its expert bloggers, then a good measure of success would be the number of people searching for, for example, [e-consultancy bloggers]. This is a success not simply because the TV campaign is bringing people to the site, but because we can see people are now beginning to associate “bloggers” with “e-consultancy”.
Catch Phrase Traffic
Some of the best adverts create memorable slogans or catch phrases. In some cases – not all, but many – it is advisable to optimise for that catch phrase. For example, if you run a radio ad with the slogan “Smile me crazy” then you may create a search frequency for the phrase. You know this search frequency is coming, your competitors do not and so you should be strongly positioned to be the organic number one for the phrase. If you can see people searching for “Smile me crazy” and finding your site then you have a measure of the success of the radio campaign. If you have strong organic rankings for the phrase but no one seems to be finding your site through it, then that’s a suggestion the catch phrase is not as catchy as initially hoped.
If you are splashing the cash to bring a famous face to a TV or poster campaign then be aware that this can also create search terms. On my blog elsewhere I wrote about Virgin Media’s adverts with Uma Thurman. As a result, it started to pick up traffic from people searching for [Uma Thurman advert] and other variants.
There are some keywords which may classify as a failure. For example, I have blogged about about Sky as well as Virgin Media. In my traffic analytics I can see people searching (and finding my blog for) [Sky Uma Thurman adverts]. Ouch. Poor Virgin Media – some people out there have watched those adverts and have thought they were a pro Sky message. I don’t see how this is possible, nevertheless those search terms are there and so we learn never to overestimate the public.
Sometimes generic keywords can count as negative. I have also blogged about the Bom Chicka Wah Wah Lynx adverts and picked up plenty of traffic as a result. This is a good example of how a catchphrase and the traffic it generates can be used as a measure of success. The blog has recorded hits for phrases like [bom chicka lynx playas] which could be considered a success as the searcher is clearly aware of the LynxPlayers site and is now trying to find it. I also have hits from phrases like [bom chicka wah wah pictures] and this particular term might be considered less of a success. This particular searcher may just be after a picture of his favourite model.
Google has a free service which looks at search frequencies rather than traffic. This can be a better macro-view of the success of an offline campaign. There are some catches. In order to be measured by Google Trends, you need to generate fairly significant amounts of search and in some cases only the most well known of brands can do this. Google Trends is historic too. You can use the service to view search trends from 2006 but not review how well you did last week.
Many analytic packages allow you to view where in the country traffic has come from. This is ideal if you are running a geographic specific campaign – such as a billboard or bus poster. For example, hiring the side of a bus in Edinburgh should result in an increase in traffic to your site from people in Edinburgh. In Google Analytics you can get to this report by clicking on “Marketing Optimization”, then “Visitor Segment Performance” and then “Geo Location” or “Geo Map Overlay”.
Google’s Webmaster Tools
There are webmaster console reports available from Google which reveal the most popular search terms which result in your snippet being shown to the searcher. This information is different than traffic analysis because these searchers do not necessarily click through to your site and become traffic.
Traffic Pit Traps
There are some pit traps to be aware of when you are reviewing traffic. Is your traffic affected by the weather? For online entertainment and even news sites, the answer is likely to be yes. Is your traffic seasonal? Travel is seasonal and finance can be too. Are you in Google News (or another search news aggregator)? If you are then traffic becomes extremely volatile as all it takes is for one story to make the “front page” and you’ll benefit from a traffic surge.
Just as with organic search, you can use paid search traffic on your brand term to measure the success of an offline campaign. More searches for your brand suggest an increased awareness. It is worth noting that this in itself is a debate. Do you want to be bidding on your brand? This blogger notes that the phrase “the long tail” has been over sold – for many sites the long tail is the place to be – but this simply is not the case for everyone. However, there are some instances when brand bidding makes sense except during a large offline campaign. These are worth working out in advance. You can sometimes coordinate with affiliates and let a cost-per-action model protect your brand spending when search frequencies might be artificially high.
Brand Hybrid Traffic
Hybrid traffic is a good gauge of success during an offline campaign. Let’s just say that you are a jobs site and a particular emphasis of your offline campaign has been to associate your brand with banking jobs. An offline campaign will have been a success for you if there is an increase in searches and clicks for “your brand + banking jobs” terms. This will also allow you to extend your exact match bidding, possibly your phrase match, and pick up these clicks more cheaply.
Catch Phrase Traffic
A clear way to decide that an offline, catchphrase based campaign has been a success is if you are able to bring converting traffic in cheaply from the catchphrase rather than an expensive term. For example, rather than paying £8 a click for [health insurance] you can bid 8p for [smile me crazy] and as long as your conversion is not 100 times worse then you will be saving money. You would have the offline campaign to thank for this.
One advantage paid search has over organic search is that you can see exactly how many times your creative was shown to someone. In organic traffic you tend to need someone to click on your result in order to measure them but this is not the case for paid search. If your CTR improves on generic terms then this is an indication that your brand association on those generic terms has also increased.
It is generally best practice to test some new paid search creatives while a significant offline campaign is running. Catchphrases, coordination and references to the offline campaign can be included in the creative display and if these result in increased clickthrough rates or, better still, improved conversion then we have another success indicator for the offline campaign.
Take a snapshot of the number of references to your brand or marketing material in Google’s Web Search and in Google’s Blog Search before your campaign. Check the figures again after your campaign. If you’ve provoked people into talking about you then these figures will have gone up. For example, Lynx’s [Bom Chicka Wah Wah] would have had no matches before the campaign started and today there are 149 blog search results and 201,000 web search matches.
The blog search engine Technorati keeps a track of which tags people are using to describe their blog posts. Large brands and successful offline campaigns can produce tags of their own. For example, Technorati is aware of 16 blog posts and 59 photographs tagged “118118”.
Flickr is a great asset for offline marketing. The picture sharing site has its own “interestingness” algorithm. Photographs – which can include poster mock-ups or photographs of the finished poster – gain high “interestingness” by being visually impressive, by being viewed quickly and often by people and by people bookmarking or making comments on the photograph. Bigmouthmedia uploaded a number of different magazine adverts and discovered that the SEO ~ PPC ~ BTW ~ WTF ~ L8R scored the most “interestingness” of the bunch. That helps us decide which advert to run the next time we have print space to fill.
Flickr can also be used to catch some of the “social media” traffic and channel it back to your site. Many blogs and forums will not link to your site but will link to your “friendly” Flickr page showing your memorable poster.
A strong – and brave – way to measure the success of a TV advert is to trial it on YouTube. A slightly less brave approach is to upload it during the campaign. Do people comment “wow” or do you pick up “sux” instead. Just be aware that a non- trivial percentage of comments on many social sites like YouTube are always negative.
Even if you disallow comments you are still able to see how people rate (out of five) the video and how often the video has been watched.
If you run an affiliate campaign then you have access to a small army of marketers. In some cases affiliates can be a good “third opinion” on whether a creative strategy is working or not. If you introduce new sky-scraper banners for your affiliates based on your current advertising campaign – and they are largely ignored then that’s a sure sign that these grassroots marketers do not rate the campaign.
Another approach is to make the campaign related banners available to the affiliates at the start of the campaign (measure the take up) and then advertise banners to the affiliates a week or so into the campaign. If, at this point, the pick up of your banners increases dramatically then you could conclude your affiliate eyes and ears have seen some of your campaign and like what they have seen so far.
Competitor analysis – for both organic and paid search – is another good way of measuring the success of your offline campaign. If you are able to provoke your competitors into adjusting their own search marketing campaigns as a result of your offline campaign then you could count this as a success. Making your competitors spend money or divert money from elsewhere is often a success.
Protecting catchphrases can be important here. If you are spending a small ransom on a TV campaign to promote a catch phrase or a new concept then you may want to stop competitors from riding on your coat tails. If I was promoting my [smile me crazy] catch phrase then I may well speak to the lawyers to see whether it is unique enough to be trademarked. If it is then I would start up the (often lengthy) process of having the search engines ban it as a paid search keyword. I may also add it to my banned list for affiliates.
Andrew Girdwood is head of search at