Marketing has never had so much data and tools to influence purchasing decisions. Yet, making the right decisions remains a challenge. Without claiming to have THE solution to this formidable challenge, we propose an approach. It consists of a clear and transparent decision-making process for all.
For several years now, the ability to collect data has grown faster than the ability to learn from it
For example, the demand for data science is much higher than the demand for the data lake. Forbes magazine goes further and talks about “infobesity” that affects most marketing teams. Data saturation affects all large companies.
There are several extra findings in this area:
Many companies are affected by the over-equipment of data collection tools. Ad servers, tag management solutions (TMS), analytics tools, marketing automation, and DMPs are all vying for the holy grail of data collection. And the need for this is quickly felt as soon as you start activating multiple media levers. The result is duplicate data collected with different methodologies. As a result, the measurement gaps become plethoric.
Media levers are becoming more diverse and complex. However, the workforce required to coordinate and operate them remains the same. There is a clear indicator of this trend. It’s the evolution of the web marketing agencies’ business model. According to an estimated number of person-days, Fees are now increasingly invoiced, and no longer on a percentage of the media buy. This, by the way, allows establishing a more virtuous partnership logic.
But strangely, the management of marketing actions has remained complex and uncertain. The chain of command is split between the marketing department, the operational staff, the service providers, and the tracking managers. Decisions are made according to one’s “feeling” or according to partial data analysis. And we apply test & learn, but each time a resource leaves, we lose part of our lessons learned and go back to square one.
So how do you manage all your marketing levers and make sure you make the right decisions? extra
A good marketing decision results from the confrontation of several “for actions” or “insights” that can be translated into concrete actions. The latter are themselves the result of data analysis and finally of our intuitions. Yes, it is high time to assume that this last component remains essential. It allows us to correct data analysis biases or to put the results into perspective if necessary, with just a little common sense.
Once this is said, it gives us the following equation:
Data analysis + Insights = Insights (for action)
Decision-making in marketing can be compared to a sea of hypotheses of actions, among which only a few are the right ones to lead to better results.
The method consists in:
- Clearly state the initial problem
- List and select the best solution hypotheses
- Test the selected solutions
- Analyze the results and make the right decisions.
We can illustrate it with the diagram below:
Ask yourself the right questions (What problem do you want to solve? For what purpose?)
If I had one hour to solve a problem, I would spend 55 minutes thinking about the question and 5 minutes thinking about the solution.
The key to solving your problems lies in the way you ask the question. Writing it gives purpose to your decision-making process. It clarifies the “why” and the final intention. It also helps define the context and the scope of the analysis.
Like the example below, a good question should identify: the challenge to be met, the time frame and the success indicator.
“How can I increase my site’s conversion rate by 1 percentage point by the end of the quarter?”
List and select solution hypotheses (How to solve the problem?)
Look for the reasons for your problem and possible solutions. To do this, put aside all basic assumptions and remain pragmatic. Each answer must be concretized by several concrete actions with a follow-up indicator to validate or invalidate the hypothesis.
For example, to answer my first example of a problem, I could decide to improve the offer’s presentation and judge the performance on the increased engagement rate on the new proposals. I mean visitors who spent time interacting on my new offer page by engagement.
Once this is done, select only those with the highest potential for success. This process implies an agreement of the participants on the selection criteria. It brings together all the referents concerned (e.g., business manager, data scientist, data analyst, tracking manager, and the acquisition/loyalty manager).
Test the selected solutions (What testing framework should I put in place?)
Define a test framework including a steering indicator, a duration, and a sufficient volume to be reached to learn from it during data analysis. Then create a test segment to compare to the rest of the traffic.
Test each solution separately to isolate their impact on performance more efficiently.
For example, suppose I want to use a new media lever. In that case, I will judge the lever on the evolution of the conversion rate over two months with a minimum volume of 1000 conversions to reach. As far as possible, I will ensure that all my other parameters remain equal. That is to say, the same advertising message and an unchanged offer.
Analyze the results and make the right decisions (What are the lessons learned?)
The objective of the analysis is to validate or invalidate the different solutions. In 80% of the cases, simple data analysis methods will be sufficient to help you in your decision-making. Advanced techniques such as predictive analytics are only used in the remaining 20% of cases.
- The two main methods are correlation analysis and trend analysis.
- Correlation analysis: This is used to identify relationships between several indicators and understand how they are related. It involves identifying patterns that highlight deviations from expected results.
- Trend analysis: This is most often used to understand changes in sales and revenue. It implies the definition of a period for the study.
Finally, based on the lessons learned from your data analysis, you will be able to make your decision. Keeping the history of your tasks learned constitutes both your know-how and the guiding thread of your actions.
In summary, marketing decision making is the result of an iterative process that includes
- The precise definition of a starting problem
- Listing and selecting the best possible solutions
- Testing the solutions
- Analyzing the results to draw lessons from them