How can you measure marketing effectiveness?
If it gets better, how can you measure the improvement?
By business modeling. The modeling process itself will uncover any holes in your marketing process.
How to Measure Anything
A “measurement” is any observation that reduces uncertainty. As I’ve written previously:
[E]verything that truly matters in business can be measured somehow.
This claim always generates push back by people who either haven’t thought their outcomes through, or who think “measure” can only ever mean “measure precisely with a meter using an internationally defined standard unit.” […]
Some best practices for devising metrics for hard-to-measure things:
- Visualize what the result of this thing is
- Decompose variables into parts
- Find others who have measured it
Let’s do that with Marketing. For example, a pizza restaurant might send out coupons via direct mail, or a CPA might offer a discount to members of her local Chamber, or a web designer might place ads in social media, or a coach might give free talks. In each case the process is similar.
We’ll use a simplified version of the approach documented in the Guide to Business Modelling by Tennant and Friend.
- Define the fundamental business question
- ID the key outputs (numbers) the model will need to produce in order to answer the question
- ID the key inputs the model will need
- Explain how the inputs get transformed into the outputs
- Write the model
- Gather data
- Present the findings
Fundamental Business Question
Too few business people give this enough time. Investing thought here will pay major dividends.
Example question: “What value am I getting from my marketing activities?”
This may be too vague. For the aforementioned coach, a better question might be, “How much money, and other valuable returns, am I getting for each dollar and hour I’m investing in giving free talks?”
Now we’re forced to turn the question into rough numbers or measures. What outputs, if I had them, would answer my question?
For this coach, we find that she enjoys giving talks — so the “other valuable returns” could be enjoyment of some sort. Note the risk — it becomes easy to rationalize doing fun things that don’t actually drive revenue. Such activities are not marketing, they are recreation. This is useful, as it reduces our uncertainty as to the effect the talks have on revenue, and thus whether they are income creating (so in a downturn we would do more of them) or purely fun (so in a downturn we might do fewer).
“Other valuable returns” could also include referrals. If our coach gave a talk that impressed Joan who referred Angie, we don’t want to lose track of the seminal power of the talk.
Enjoyment — on a scale of 1 to 10 — is one output. Another is revenue brought in by leads generated by the talks.
Next we have to think about what inputs might exist, or need to exist. It’s common to discover that you’re not tracking the data you need.
The outputs and the business question imply some of the inputs — such as how many hours the coach spends preparing and giving each talk, and how much money she spends on handouts and travel.
For revenue generated by talks, we verify that the leads are tracked by how they came into the system, such as from a talk. I recommend that a talk on Goal Setting to a Rotary group on December 14, 2015 be coded in a ‘lead source’ field as “151214 Rotary Goals” — this is short, it starts with the date, and the date is formatted YYMMDD. That allows us to alphabetically sort the ‘lead source’ field and the result is in date order.
(If you don’t have a place to track leads, I recommend starting with Zoho CRM’s free version. I’m not compensated for saying that, though maybe I should be. Be sure if you use GMail to sign up for Zoho using Google Authentication — it makes for a better experience.)
For tracking revenue from referrals generated by talks, I would just use an entry in the Notes field for Angie: “Referred by Joan (via a talk).” Looking up Joan would show which talk. As a coach will not have very many unique clients per year, this is sufficient. For another case where data volumes were much higher, we would use more sophisticated referral and lead source tracking.
For tracking the coach’s enjoyment of her talks, I suggest that she fill out a short report card after each talk, indicating how many attended, whether it was enjoyable (on our agreed scale of 1 to 10), and also what aspects she enjoyed most. These report cards can be paper or electronic — whatever is easiest. Even a very prolific speaker will not have more than 100 talks in a year. (A simple method for gathering such data is via a Google Form.)
Explain How the Inputs get Transformed
For a sophisticated model, we would want to explain any interesting calculations. For “Lifetime value of a customer” I would explain which formula I used (the easiest I’ve found is the simplified model here).
Any business model will benefit from outside ideas. Anything you measure, has been measured by others. For our Marketing model, I might scan the work of the Marketing Accountability Standards Board (subscription required for some pages), to see what subtle “gotchas” have already been learned by others, and to look up standard definitions of key terms.
These explanations can be written as text in the spreadsheet (see below) or in a separate planning document.
Write the Model
With the above work done, this step is simple.
Far too many spreadsheets cannot be used by anyone other than their creator, and often the creator forgets their function after leaving it for a time. To counter this, use this guidance:
- Place all inputs in an INPUTS tab
- Place all calculations in a CALCS tab
- Place all outputs in an OUTPUTS tab
Only the Output tab needs to be formatted for display or printing.
(Download the sample spreadsheet in Excel format here.)
Now you get to gather the data you need and plug it into the model. If the data is currently lacking — as is often the case — revise your systems to start collecting it.
For our coach, we create the aforementioned scorecard to track the talks, we look at the sales system to verify that it tracks ‘lead source’ and ‘referral source’ accurately, and we check it periodically to make sure it doesn’t drift off into inaccuracy or incompleteness. We revise the coach’s list of ‘lead source’ values to include which specific talk led to which lead.
A nice bonus of using a CRM is that it becomes easy for the system to give you interesting data. For a lead who gave his card to the coach at the end of the talk, his ‘created date’ in the system is probably the next day after the talk. For another lead who emails a week later, his ‘created date’ shows a lag. The dates of the eventual sales will also reveal how long it takes leads to turn into sales.
Marketing is hardly simple.
If someone encountered three of your marketing activities and finally contacts you from the accumulated impact of all of them, you’ll probably give too much credit to the latest one.
(What you choose to measure and call ‘success’ matters. Should marketing just be measured by leads, or by sales? In one famous example, Sheena Iyengar set out samples of jam in a supermarket — in two trials. Which was better? The first trial outperformed the second by 3:2 based on how many people stopped and tasted jam. But the second trial outperformed the first by 6:1 based on how many people purchased jam that day. If you only measured tastings you would do more of the first approach; if you measured sales you’d switch to the second approach.)
The purpose here is not to be perfect, but to start measuring.
One last warning — most CEOs insist that they’re tracking lead sources. But audits show that in fact they aren’t. Don’t just assume — go look.
Measure Marketing Effectiveness
Ultimately it’s your job as CEO or owner to ensure your marketing is effective. Whether you do this yourself or delegate it, you still need to be confident that it’s being done coherently and methodically.