Discontinuities and Corruption

- and putting the goal-post where the ball is

Often when the regime changes, specifically for a company, there is supposed to be a process of handover. If there isn't much of that, it's assumed to be a time to start with a clean slate. What is interesting is that it's also a time for the most common and the easiest kind of corruption - being insincere with your work.

Don't get me wrong, I'm not saying humans are corrupt when they get a chance - I'm only, for now, saying that it is a possibility. Humans are known to optimize.

When you take over a new position at a new company, you'll rely on the existing employees, juniors and peers, to give you background. Some times there are too many changes in a short while in companies. Sometimes a few exits snowball into an avalanche. It is common to assume the ship is sinking when mice are seen jumping ship. And at those rare occasions, legacy is in the custody of very few people. Those few people can change the history if they want. It gets even more complex when the previous leadership is not contactable or hostile and no longer interested in your welfare.

What that leads to is an interesting phenomenon that I have heard mentioned across companies. The goal post can move now - to where the ball ends up. If profits are down and turnovers are high, you'd be told that's exactly what we set out to do. If an irrational decision was taken, it was taken by the previous leadership. People will come to you mentioning past promises made to them of promotions and raises.

Most new leaders pass - there's so much more to be done in the initial few months that the archaeology can wait. Some dig in only to get trapped in the mud of conflicting versions and unclear evidence. No one wants to start off doubting peers and juniors. I'm still wondering what is the best thing to do, and how....

Piloting Non-existent Concepts

One of the more interesting discussions I have had around Strategy is around Pilots. We all know that in general, and specifically in emerging or nascent markets, is that there isn't enough past data to base decisions on or to make business plans on. What you do in such cases is not research but pilots? Bet small amounts on multiple horses, see which horses win their first races, bet more on them and less on the others and so on till you have a winner.

Talking of Pilots, it is possible in eCommerce, more so in eCommerce than in offline commerce, to pilot out hypotheses. If you have a great packaging solution, try and see how the product does without the fancy packaging but discounted to that effect, check how the same listing does with or without CoD, or with or without no-questions-asked returns. If website / platform flexibility is an issue, one could try out the demand pilots purely on mailers. If one wants to know what's the sweet spot on prices or discounts, or the tradeoff between, say, faster delivery and price, it is possible to have two listings at different prices and different delivery timelines to check which one takes off faster in sales. It is then possible to dynamically alter the discount and the delivery timeline to check what's the point at which decreasing returns set in. And this brings us to the crux of this discussion.

How does one pilot faster deliveries - or any other offering we don't have and that takes sunk investments to build-out? We already operate on optimal delivery schedules. Won't we have to set up a separate infrastructure to create faster deliveries? And if we have to invest in setting up stuff, then doesn't it defeat the purpose of piloting the concept? We can't create a warehousing and pre-shipment infra and then conclude that faster delivery doesn't create value in the eyes of the consumer i.e. the consumer is not willing to pay more to get the goods faster. Then we can't roll back our investments to go back to what we had earlier.

Or maybe we can.

We could partner with a third party to create this experience for the pilot, but then the costs of doing this won't be representative. But there's something simpler we could do - which is not exactly the same thing, but quite useful for pilot results.

We can pilot slower delivery!

Now we'll put up two listings side by side, one with our standard price and delivery-time, and the other with a lower price but slower delivery, and check which one appeals to consumers. Versus a listing with an elevated price and faster delivery, this simulation will also give us results on what is the sensitivity on between delivery-time and price.

You see? There's always a way. When the ideal option is not possible to pick, pick the best possible option.

What are your thoughts?

Where is the Devil in Business Plans ....in the Assumptions?

...is in the details, right? Well if you know where to look. I have often been asked what I look for when I check a business plan. Many people I know spend a lot of time going through the data, the data-sources, the formulae, the methodology (e.g. discounted cash flow valuation) etc. Some even check for links between data-sheets, references and so on. Where I spend 90% of my time, based on my experiences so far, is on the assumptions sheet.

I'm not saying, for a moment, that the output sheet is not sensitive to the other things highlighted above. I'm saying something else. Most people who build business plans can be trusted for accuracy on formulae and linking of cells. Where competence has a smaller role to play is on the assumptions. Assumptions are futuristic, they are usually based on data but there is a choice of data to choose from e.g. for the GDP or inflation data, one could have different sources, all authentic, saying different things. Which source you pick determines which data point is picked up and how your business plan results look. A lot of assumptions are not even data based e.g. assumptions of how much market share one will get in year-3. There is no sure-shot way of getting that data. All you have in such cases are scenarios.

Next interesting point, how many business plans have scenarios built up? Even if it's not a complex monte-carlo simulation here, just basic scenarios - base case, pessimistic, optimistic etc. Now when you apply scenarios on the end-results, which is how it's commonly done, you'll end up pushing up results by 15% for one scenario and down by 10% in another - but that means little. What one could and should do is apply those to the assumptions sheet. Pick up the lowest GDP growth estimate in the pessimistic case, the most probable estimate in the base case and the best estimate in the optimistic case - and so on for all the assumed variables. This gives you a much wider range of what is the worst that might happen when all goes downhill and what is the best case when all looks up.

Now let's take yet another perspective, of biases and vested interests. Often the person who creates the business plan is the person who stands to gain or lose the most by the overall verdict on the plan. If this person were to have an ulterior motive, where do you think he / she plants his / her biases? It'll probably not be in the formulae or methodology - since getting caught there could be the end of that career. It's more likely to be in minor tweaks of assumptions one could easily reason or argue out of.

One more perspective is on what happens to business plans over time. The calculations don't change, what changes is all on the assumptions sheet. Interest rates change, market or competition developments happen, new products hit the market, and the impact is again on the assumptions sheet, and still when we look at our follow-up processes, so very often we tend to focus on the summary sheets e.g. RFP vs. actuals, projections vs. latest forecast etc, rather than setting up a frequency where all the past business plans are reopened and the assumptions revised.

Either way we look at it, checking assumptions is most critical. Updating assumptions periodically is also very important. Finally scenarios - do not accept business plans without scenarios, and without those scenarios being linked back to the assumption-sheet where - you guessed it - each assumption has scenarios.

What has been your experience? Have you caught more devils elsewhere? Or is your experience the same?