Tracking the offline world using online tools
By now I'd hope that most marketers are doing some level of tracking for their online marketing efforts. Even if you are just measuring the basic front-end metrics of impressions and clicks, at least its something (though you really should focus on the business outcomes, not the front-end metrics). For whatever reason, marketers have traditionally shied away from tracking their offline tactics. Quite honestly, its been an arduous process to collect data and most marketers lack the analytical prowess to make correlations. As a result, most marketers don't understand the effect those tactics are having on their business and offline tactics become a gut-feel exercise. In recent years, one of the reasons advertising is moving online is the enormous trackability of it.
This post aims to give marketers a few ideas on how to track their offline efforts using online tools. Since most people use the internet in some fashion before purchasing something, even an offline transaction, there are ways we can utilize the web to analyze campaign effectiveness. Like most of my posts, I am using Google Analytics for the examples but you can do this with whatever analytics tool you have like Omniture, Coremetrics, Webtrends, etc. Taking it a step further, Omniture has some pretty wild tools like Omniture Insight that can tie in multi-channel data (such as retail sales) into one place for analysis. For this post, we'll keep it simple....and generally free.
First, let's talk about the mechanics of how you'd track your offline tactics.There are three main ways you can track the effectiveness of offline efforts using online tools.
- Use redirects to append campaign tracking codes that can be tracked in your web analytics solution
- Look for changes in behaviors, such as uplifts
- Redemption offers
For those not yet savy on redirects, here is the low down on how they work. Basically, you can create a pretty URL (known as a vanity URL) that is easy to remember so when a person sees/hears your ad they can easily remember it and type in a web browser later. When the visitor types in the URL it actually redirects the visitor to another landing page URL. Vanity URLs technically don't exist, which has caused me problems over the years. I've had marketing folks run campaigns with a vanity URL and then after the fact ask me how it did. The problem is since the vanity URL isn't truly a page it doesn't fire the web analytics scripts to track it. Because vanity URLs are typically typed in there is no referring source either, so it looks like people just showed up magically on the site. As a result, no tracking back to the vanity URL.
Alas, there is a solution to that problem but takes some planning before you launch your campaign. The workaround is in the redirect being used to drive the visitor to the landing page. This type of solution works for TV, print, events, radio and anything else offline. Here is an example:
Say I create a print campaign for a brochure that I am passing out at the next Triangle AMA event and I want to track the impact of it. On the brochure I have 2 calls to action:
- A phone number (more on this in a bit)
- A vanity URL that goes to capstrat.com/analytics
My hope is that people will read the brochure and thinks its so overwhelmingly awesome that they decide to contact me about what we can do for them with regards to analytics. Before the event, I've made friends with our IT guy and set up a redirect that takes people from capstrat.com/analytics to a URL with campaign tracking codes like this: http://www.capstrat.com/?utm_source=triangleama&utm_medium=brochure&utm_campaign=hazen_octobercampaign
When a visitor goes to their web browser to type in capstrat.com/analytics I know that they came from that brochure and can tie it to whatever outcomes happen on the site, such as filling out a lead form, buying stuff, reading blog posts or downloading whitepapers.
Important thing to know - you can only use a one vanity URL to one redirect. So I wouldn't want to use the same vanity for my national radio ad that I am doing. Instead I would use a different unique vanity like: capstrat.com/radio which would redirect to a URL with campaign tracking codes: such as: http://www.capstrat.com/?utm_source=national&utm_medium=radio&utm_campaign=hazen_octobercampaign
Inside Google Analytics I would see both tactics and be able to figure out if they were effective or not, as well as compare it to my online tactics.
There are a a few redirect possibilities that are expertly explained at Brian Clifton's blog (download the whitepaper to check it out).
In the brochure example above, I mentioned the phone number. There are some other clever ways to track that. I could utilize something like GoogleVoice to create a unique phone number just for that piece of collateral. The phone number would redirect to my normal phone number, but the benefit is I can track the unique phone number. When I get leads from that phone number I would drop the source of the lead (in this case the brochure) into my CRM system and track it through the opportunity management process. There are more sophisticated ways to do this as well but you should always consider using unique phone numbers to track effectiveness.
While the first example looks at utilizing direct attribution methods through campaign tracking codes, the second way to evaluate offline tactics is less of an exact science. The second method involves using changes in behavior based on isolation whether its geography or time.
Still using my brochure example, let's say I saw a big spike in visits going to analytics related content on the Capstrat website a day or two after the brochure was unleashed at the Triangle AMA event. I can use isolation to determine if there is a strong correlation to that jump based on the timing of the spike as well as the geography of the visitors to the site, since most of the attendees are from the Triangle area. I could look at other site behaviors such as downloads of my presentation, internal search keyword changes, people contacting me from the website as other correlation points to the success of that brochure.
I am a big fan of doing geographic isolation as well to determine uplift. Example - I could run print ads in Raleigh but not Charlotte and do everything else the same. I could measure uplifts over the expected baseline in Raleigh to understand the impact of print ads. I am simplifying it here but to do a truly awesome job you'll want to apply some regression analysis to understand true impact.
Here are some things to look for when tracking offline tactics:
- Changes in content consumption related to your offline event
- Changes in internal search keywords
- Changes in sources of traffic to your site. If the visitors don't use your vanity URL, they might use Google to get your site or just type in your normal domain.
- Change in external keywords coming to your site
- Uplift in Google based of people searching for your company name or ad-related keywords
- Uplift in New Visitors coming to your site
- Ask Visitors how they heard about you via surveys - Did it change?
- Changes in conversions such as increase in leads, sales, etc.
- Did your share of traffic change in relation to your competitors (check out Compete, Hitwise, etc)
With all of these correlations it is important to take into account timing, seasonality, intangibles and geographic differences.
The third way to track offline campaigns is by utilizing redemption efforts. Redemption can be anything like a coupon, rebate or code that has to be used to unlock something or get a special offer. Some example cases:
- A print ad has a promotion code in it to save 10% by going to the website and typing in the coupon code when making a purchase
- A retail store gives a rebate code where you have to go to the website and will out some information in order to get the rebate. This happens a lot with warranties in order to get your name in their databases.
- A magazine ad has a code to be used to unlock protected content on the website.
In the old days, redemption was a lot cleaner but in recent years has become less accurate from a measurement perspective due to the rise of coupon sites. The coupon sites post the same promo code on their site (literally within minutes of a launch) which muddies up the effectiveness of the print ad since you now have multiple places where the code can be found.
So there you have it, some fairly easy ways to measure offline events. If anyone out there has other ideas or questions, feel free to post here. To hear more about this topic, I'll be speaking in more depth about it at the upcoming Triangle AMA Analytics event.