First of all – credit to the man who pointed me to this important document Hugh McEvoy, Head of Product at Sefaira, where I also work.
Secondly, here is the link.
Thirdly, I sped read it so will be coming back to update but here are my first impressions in a self contained 30 mins;
- The journey can not be completed in one step so be humble enough to accept why people cannot see the finish line and use the steps that are already there to bring people along. You do not need to build a whole new road to a new destination if there are flagstones laid already.
- Customer development is time critical and can work both ways, being quick on the up and quick on the way down.
- Predictions are hard – hybrids are great and found a market but at around 3% of new sales they are not the breakthrough predicted to be 35% of the US car market. So the theories in the book/paper are as susceptible to the external market and do not produce a product which is intrinsically better but simply help to avoid massive failure.
- The split between senior management and sales/marketing is an anachronism and, I think that turning this on its head is a key to success – efficiently using the data points garnered by the whole team to crowd source priorities. Everyone has their different motivations and biases but also their own skillset. It is a rare senior manager who can do everything well. So, if they are trying with limited resources to deliver feature improvements they may actually not be best placed to get the head up and look at the next feature/product.
- A different model to the Product Development Model is smart but the baby does not go out with the bathwater. You still have to finish the feature the customer tells you it wants, whether it is called Beta or MVP, before you can draw a conclusion.
- Customer Discovery is an essential way to think about matching and targeting customers ahead of having a product/feature. Do not extrapolate from the early adopters, especially if you are one yourself.
- PERT analysis (Program Evaluation Review Technique) uses probabilistic time estimates for the program definition
Optimistic Time + (4*Most Likely Time) + Pessimistic Time
- There may be a lot of confusion around whether you are a new product in an existing market, a new product in a new market or a new product in an existing market which you are trying to re-segment as a niche entrant or low cost entrant. This is a key difference and should be defined and debated and redefined before action.
- A Bass Model helps forecast adoption rates for “new-to-world” models (based on an S curve but distinguishing between early adopters and more conservative adopters) using coefficients of Innovation and Imitation. We either need a similar product to extrapolate from or to extrapolate from several data points – otherwise it is pure speculation and we are likely to weigh to heavily to early adopters.
- The Technology Life Cycle Adoption Curve and The Chasm ARE REAL THINGS! Do not diminish them, in the same way as dislocating the sales and product timeline, someone should be looking at the chasm even in the middle of all the early customer problems.
- Early adopters should be seen as problems and treated as such.
- Learning and Discovery cycles, although essential, should not be seen as a virtuous thing in themselves but a means to an end. Do not re-learn the same thing twice and again. Be humble and honest the first time.
Probably the most important point would be to align philosophies like this as a senior team at the outset and reaffirm through each key step for the company. The customer validation process could so easily be misunderstood. The acceptable minimum in an MVP could be different in so many heads. The point at which it is assumed scale can be achieved should be honestly revised most often as this step, above all steps, changes the drive of everyone and sales must be able to sell what they have.