Friday, April 22, 2011

Lesson #22: How to Calculate ROI on Your Marketing Spend

The worst thing an entrepreneur can do is dump a bunch of money into marketing spend, without having the appropriate reporting in place to track the results of such spend.  Although marketing can often feel like a "creative" area of the business, which it is, it is more importantly, one that requires religious "financial" discipline.  In this post, we will discuss how you infuse ROI disciplines into your marketing DNA and team.

Let's start with internet marketing, one of the most trackable mediums in the history of marketing.  Based on ad tracking tags, web server logs and other software, you can track each ad impression, and the resulting clicks, leads and transactions therefrom.  Whether that is search engine marketing, or email marketing, or social media efforts, or affiliate partnerships with third party websites, the data is easily captured.  But, the key is to make sure you are using these ad tracking tools and reporting on each of these inputs.

While I was at iExplore, my marketing dashboard told me data like: 1MM visitors came to my site that month, sourced 50% from search engines and 50% from affiliate partner sites, resulting in 1,000 people contacting our call center to book a trip and 200 actual transctions.  And, more importantly, I could disect my search engine performance between Google, Yahoo and Bing, and even further, between paid search and free search efforts within each.  I also knew that my affiliates like National Geographic were driving more leads and sales than affiliates like Travelocity.  This was very valuable data, if studied and used, to dial up and down my marketing spend online, moving monies from underperforming vehicles to over performing vehicles, to maximize my marketing ROI.

Although it is harder to do, you can also use the same level of tracking on your offline marketing activities.  As an example, if you send a direct mailer, use a unique promotion code that the customer shares at the time of purchase with your call center staff.  If you are promoting a 1-800 number in your TV advertisements, use a unique 1-800 number for each cable channel you are advertising on, or more specifically, each specific program you are advertising on.  In your magazine buys, use different promotional URLs to point your traffic by magazine (e.g., .com/TimeMagOffer). Not all consumers will end up contacting you via these tracking mechanisms, going to your main website instead.  But, a good portion will, and you will be able to make smart interpretations and extrapolations from there.

The overriding point, where you can, is to track each distinct marketing effort on its own stand alone merits, to assess whether or not you want to continue investing in that vehicle going forward.  And, worth mentioning, never start spending your full budget day one.  Take 10% of your budget, test a bunch of different vehicles you are considering, and then invest the remaining 90% of your budget in the best performing vehicles from that initial test.

Then, once you know the profitable metrics for your business (e.g., never spend more than a $2CPM on any online banner ads), you need to religiously live by those metrics.  Only buy ads that fall below that criteria.  And, often times, your desired sites will not allow you to buy committed advertising at a level well below their rate card.  So, in those cases, you need to be really creative when negotiating those deals.  Tell them you don't need to buy committed space at $10CPM, you are willing to buy remnant, unsold inventory at $2CPM, which may be more digestible to them instead of letting remnant inventory going unsold for zero revenues.

If you cannot get the appropriate line-by-line tracking of your marketing investment, then you need to make sure you are studying the overall impact on your business from that aggregate marketing investment.  For example, if you put an additional $1MM in marketing dollars to work in January, what was the resulting incremental lift in overall revenues in the following months and did the incremental profits justify the investment.  But, when you do this, it is critical you are comparing apples-to-apples data.  For example, make sure seasonality isn't skewing your data (e.g., a retailer's big Christmas sales season naturally took up revenues, not the increased marketing). 

And, it is equally important you know your normal sales cycles between the time marketing is placed and the time a transaction actually closes.  For iExplore, it was a three month sales window, and for MediaRecall it was a 6-12 month window (so study the lift in revenues in the appropriate sales window for your business).  But, for long sales cycle businesses, you do not want to keep your marketing running for too long without knowing whether or not it is profitably working.  So, instead of waiting for actual closed revenue data, study inbound sales leads immediately after the marketing period, to ensure the leads pipeline is building up fast enough to justify the marketing investment (based on your estimated and historical sales conversion rates).

Marketing departments are typically run by "creative" types, that don't necessarily understand the "financial" side of the business.  It is up to you, to make sure these financial disciplines are followed to avoid potentially wasting a lot of mis-spent marketing dollars.

For future posts, please follow me at:  www.twitter.com/georgedeeb

2 comments:

Aashik said...

Nice one. Thank you :)

How do you think can we approach to find the ROI on outdoor advertising?

George Deeb said...

Outdoor is harder, but doable. You will either need to ask clients how they first learned about you, at time they buy with you, or via email surveys post purchase. Or, promote targeted URLs or phone numbers, specifically tracking the outdoor activities via those links or calls. Hope this helps.