Data Driven Selling: Good Moves to Make and Traps to Avoid
“Above all else show the data” reminded Edward Tufte, American statistician and professor emeritus of political science, statistics, and computer science at Yale University, often hailed as the da Vinci of data. This time around, Idea Watch picks up things you must do more of while engaging in data driven selling and traps you must avoid
Four good moves to make when making Data Driven Decisions:
Here are four tips from “Data Story” by Nany Duarte, Communication Expert, Stanford Guest Faculty and Author of The HBR Guide to Persuasive Presentations, HBR’s popular guidebook. While Duarte’s tips are for presentations to C Level managers in general, they are adapted here for sales presentations
- 1. To persuade managers, prove your point
Ensure that your presentation is thorough and every aspect of your proposal is supported by data points. C Level executives who sign off on all decisions, including purchases, stake their reputations when they give their approvals. So, they need to see that you have done your homework and the data marshalled covers all potential concerns and benefits
- 2. Let your data get to the point fast
Executives are known to be chronically time starved. So, skip any effort to elaborately build up a case using large amounts of contextual data. They are usually well briefed on what they need to decide on. Let your data get to the point straight leading to a brief, logical and rigorous recommendation
- 3. Respect their time and communication preferences
C Level managers have competing demands on their time. They must drive a strategic agenda and ensure that all their stakeholders are happy. That makes time their most precious resource. Little wonder they appreciate professionals who communicate clearly and are easy to understand. This calls for presenting just the most impactful data required for a decision, choosing a data presentation format that offers maximum clarity
Five Traps to Avoid When Making Data Driven Decisions
There are 5 commonly observed traps encountered in making data driven decisions. The first three are highlighted by Megan MacGarvie and Kristina McElheran writing in Harvard Business Review in their article Data Analytics: From Bias to Better Decisions. The other two, often observed in business conversations, find a mention in economics and statistics literature.
-
Confirmation Trap - More attention paid to findings aligning with prior beliefs, ignoring other facts and patterns in the data
Example: "South Indian snacks are not very well known in Northern states and hence they may not sell in those markets," pointing to some cases of past failures of South Indian restaurants in Northern states
-
Overconfidence Trap - Decision maker too sure of outcome based on past successes. Assuming accuracy of our judgments or probability of success stronger than data would suggest
Example: "Cereals are a popular breakfast food, world over. So, India, with its vast and deep consumer base, should naturally be a good market for cereals" This assumption of MNCs faced resistance as Indian consumers were not enthusiastic about soggy flakes in hot milk with a bland taste
-
Force-Fitting Trap - Data leads to surprisingly counter intuitive conclusions. Is it a signal or a neat explanation of noise through force-fitting?
Example: Do It Yourself kits for minor home repairs and home appliances for dusting and dishwashing have generally found it difficult to break into Indian markets where affordable human help and support are readily accessible for such activities. Yet signals from data and insights gathered in narrow urban pockets may be aligned to global trends. In such a case, force-fitting those insights to the national market as a whole could be a serious mistake
-
Confusing Correlation with Causation - Data may show two unconnected variables moving in tandem. It doesn't mean one drives the other
Example: For a given period, data shows that when retail FMCG sales goes up, auto insurance claims also increase. Are these connected? Does one cause the other? That may not be the case. Therefore, such movements of data in tandem need to be studied in depth before concluding that one causes the other
-
Mistaking Anecdotes for Broad Trends - "One swallow doesn't make a summer" goes the proverb. Stray instances do not constitute a trend. Anecdotes and episodes we encounter cannot be a basis for concluding that there is a pattern or trend behind them
Example: A recurring theme for discussions in sales review meetings is the perceived limitations in product range acting as a dampener on sales. The argument offered in support of this runs like this – "We see a lot of people among our friends and relatives owning X product launched by competition. We don't have that variant in our product range. That is limiting our sales." This may be a classic case of trying to spot a trend based purely on anecdotal evidence
“The fundamental task in data analysis is to make smart comparisons—we’re always trying to answer the question ‘Compared with what?’”
– Edward Tufte, author, professor, and data visualization expert
