Have you spent days and weeks working on a campaign only to find that it has lesser and lesser impact? As marketers, we often need to refocus to escape from this frustrating cycle. That’s the challenge when you are optimizing for local maxima vs. solving for global maxima in marketing.
Before we deep dive into today’s topic, let’s internalize this lesson with a story. You are the social media lead for a global consumer goods brand. Every year you manage two large global campaigns to drive awareness for your brand’s products & services.
Both the campaigns combined generate at least 30 – 40 million viewers. If you were to run an ad exposure survey, it would prove that your audience can now recall your ad. But the larger question to ask is how many products have been sold as a result of the campaign?
This is a challenge most marketers face today. We optimize for local maxima while we are supposed to optimize for global maxima of success.
Local maxima view of success: The social media team can claim they did a successful campaign using the reach generated and the brand lift as a result of the campaign. They will demand a more significant share of the marketing budget next financial year.
Global maxima view of success: The brand hasn’t generated any sales.
Sadly, most brands today focus on local maxima of success. The problem is most of us don’t think critically about what problem we are trying to solve. The company’s incentive structure also somehow rewards the local maxima.
Let me give you another example of local maxima obsession that I have seen in recent years.
Local – Global Maxima: Example: Air Freshener Brand
A large multinational launches a new brand of air fresheners and launches a YouTube campaign to create awareness for the brand. A short television commercial is developed to announce the launch.
The users on YouTube are repeatedly shown the same ad (because the agency managing the campaign probably missed setting the maximum number of views per user to one). The campaign is live for six months, and the campaign report for the year shows millions of views for the brand on the platform.
While the multinational struggles to drive sales for its air fresheners in the market.
Solving for Global Maxima of Success
So if you have you were to manage both of the above brands, what should you have measured to optimize for global maxima?
For starters, you should have looked at all the possible data points at the end of the campaign. To conclude, that campaign didn’t work for the brand. Maybe an enhanced survey that asked the viewers how likely they are to consider products from the brand over that from competitors. That’s the key intent or behavior that you are trying to change.
That would have clearly proved whether the campaign had created an impact in the market.
Time series correlations would have been a great start right after the first week of the campaign. How many people were visiting the website during the campaign? Was there a noticeable change in the number of visitors on the website or footfalls in the retail store?
If your brand is available with a large format retail store, you could have partnered with one of those stores to measure the overall impact. This data would have clearly shown you that the campaign might not reach global maxima in a few months.
If your brand doesn’t have this kind of ubiquity, then a match-market test could have been a great way to discern whether the specific marketing strategy had a business impact.
Now lets me play the devil advocate and briefly touch upon one of the excuses that you often hear that is long term efforts vs. short terms efforts. Brand building requires a longer horizon to see measurable results.
There is a kernel of truth there, brand building takes time. But to obsessively measure more than the local maxima to discern signals in the short term that illustrate that the long term brand building play is not just an excuse to flush a lot of money.
Even long term brand play is about taking away small portions of market share over time.
Let me end with a final thought, what if the social media campaign in the first example was really successful. But the brand had multiple campaigns running parallelly. How would you know which campaign helped in generating the maximum impact?
The answer to that question is to use data-driven modelling to measure the impact of online campaigns to understand cross-marketing value. Then move to matched-market tests to understand your overall campaign performance (Online + Offline). And finally, stop at Market Mix Modelling, which is more to measure the overall impact of the strategy.
