Turn your hypotheses into a syntesis

Turning analysis into an answer is an active step, you can drive.

Here is the test I applied to every piece of work that crossed my desk. Can somebody look at the output of your analysis for ten seconds and understand the insight? If not, the analysis is not done yet.

I often saw associates who had done genuinely good work and presented findings with no clear “so what.” The numbers were right, the chart was clean, and the reader had to do the thinking themselves. That is not a presentation problem. It is an unfinished analysis.

Ask:

1.  What does the result mean? Is it higher or lower than expected? Is there a trend worth noticing?

2.  How does this change or support the answer? Does it confirm the hypothesis, contradict it, or introduce a new angle?

3.  What are the implications? For the company, for the market, for the other workstreams on the case?

4.  How does this drive the client to action? Given these results, what should they do differently?

These four questions give you the tagline, the comments and the key takeaway. They are the bridge between a chart and a recommendation, and they take about five minutes.

If you are unsure whether it landed, show the output to a peer and watch. If they get the “so what” immediately, you are done. If they ask what they are looking at, you are not.

The rhythm: dive, surface, dive

There is a rhythm to problem solving. You form a hypothesis, dive down to gather data, come back up to evaluate what it means, then dive again.

Every few hours, consciously step back and check: am I still on track? Has what I found changed the direction of my work?

Many associates fall into the trap of staying down too long. They get absorbed in the data, run additional cuts, explore interesting patterns, and surface three days later with a mountain of analyses and no story. So synthesize regularly rather than once at the end.

The rhythm also protects your time in a second way. Before you dive into a detailed analysis, think about the decision it needs to inform. How much data do you actually need? Can you ballpark the result first? If a rough estimate already shows the answer is clearly off, do not go into the specifics. Save the hours for analyses that could actually change the recommendation.

Checks to do before you present

Do a sense-check. Do the results make sense in the real world? Are they realistic and achievable? Is the number high or low relative to what you would expect? Are the trends causal, and causal in the right direction? A number that looks fine in a spreadsheet can be absurd once you translate it into physical reality. If your analysis says a factory can double output with no additional investment, that should raise a flag before anyone else raises it for you.

Triangulate. Do not rely on a single source or method to validate an important finding. Cross-check against at least one independent data point, which can be the back-of-the-envelope number you started with. If your model says the market grows at 12% and three industry reports say 6 to 8%, you need to understand the gap before presenting. Triangulation does not mean your number is wrong. It means you must be able to explain why it differs.

Know your sources. For every number on your slide: where does this come from, is the source credible, what assumptions went in, when was it last updated? This is not optional. If a partner or client challenges a number and you cannot trace it, your credibility takes a hit that is hard to recover from. As a rule, I annotated every key number in my models with a brief source note, which forced me to verify before committing it to a slide.

Do not speak up without a data-driven justification. If you say something in a meeting, you should be able to back it up, or clearly label it: “this is a hypothesis.”

Gauge the focus. What is the one thing that is most important to communicate? What question are you most likely to get? Having that clarity going in makes the conversation sharper and stops you being caught off guard on your own analysis.

START THIS WEEK

1.  Make sure to note every source, also contradicting ones, into your notes or models.

2.  Make sure to pull up at least once a day to evaluate your analysis in the broader context of your work.

3.  Picture all your results in the real world. What would it actually take (People, money, space, etc.) to make these results true?

The synthesis is the result of problem solving, and it is a step in itself.