A recent experience* provoked me to write this post – when we had to roll back a product update just after a week. To handle the situation better next time, I came through “reversible vs irreversible” decision concept and developed my framework to take hard decisions and still being customer-obsessed.
This is not new. I have often found myself at these difficult spots – wanting to deserve the customer’s business and providing enough product value. On one hand, there are customer problems and on the other, there are constraints of time, resources and business.
Typically, there is data and evidence to make decisions and also know the tradeoffs. However, when we don’t have enough information or something has to be sacrificed, how do we make a decision, let alone knowing what is the right one?
Here is a framework that I think would be useful based on my experiences, but first let us understand the basics.
What is customer obsession?
It is knowing the customers and their problems really well. And then being obsessed with solving them by taking them into account for every discussion and every decision.
Initially, I used to confuse customer obsession with customer love. Why can’t we provide all these features and experiences while we are still profitable? It took some time to understand why we can’t. It is not just business constraints but also our responsibility to satisfy more and more customers in a timely and efficient manner.
Here is tactical advice to be customer obsessed:
- Understanding who the customers are by talking to them one-on-one, developing empathy.
- Stepping into their shoes and experiencing their usage. Best if we can imitate the same environment.
- Using data to develop multiple customer behavior hypotheses.
- Finding not median and average customer, but typical customers.
- Accounting customer’s comments not just upvotes while prioritizing features. Note that customers are not always right. More on it later.
- Charting out the entire journey end-to-end for small feature developments and product tweaks.
- Validating assumptions at each major step of development.
- Following constantly and iterating on customer’s problems.
- Tracking and emphasizing customer satisfaction and usage metrics more than business matrics.
- And lastly, using customer obsession in meetings even for trivial decisions by talking, explaining, and presenting in the form of customer stories.
What are ‘hard’ decisions?
When the stakes are high and there is no coming back, then we have decisions that are hard to take.
At the cost of repeating myself which has been said millions of times – Product managers constantly make decisions. In our daily sprints, inter-team meetings, intra-team meetings, planning meetings, grooming meetings and review meetings, we make decisions, sometimes even without realizing.
For most of these discussions, we would have done our homework – researched or have talked to stakeholders. For the rest, we would need more time to arrive at a consensus.
Therefore, after some time and deliberation, we would eventually arrive at decisions. And at that point, we would be comfortable for all the consideration we had. However, when there is less data, no research, limited stakeholders involvement, high impact and time is running out, these decisions become hard to take.
Buying time to alleviate hard decisions
It might seem we can get gather more data, talk to more people, have more meetings to transform hard decisions into comfortable ones. However, often in reality that’s not possible.
Buying time to research more and support the decisions will increase opportunity costs as well as demotivates everyone involved. It is a kind of product meta – People dealing with problems want to reach conclusions faster so they can create products that solve customer’s problems fast.
Just like product managers keep making decisions, they are always pressed for time. Also, consider the alternative – If the team is blocked because of a decision, which could be easily reversed, then should product managers buy time to gather data, doing discussions and then making the decision? If we are prepared for the sunken cost, work gets done much faster.
Undoing hard decisions
One of my engineering managers used to say – “We can always undo decisions. Everything can be redeveloped.” Though I agree with his spirit to move forward, there is more to undoing decisions that meet the eye.
On the engineering side – Yes, we can re-develop the feature differently. Yes, we can launch first and come back if it doesn’t work, making a different decision this time. Yes, we can reverse all decisions and come back. However, on the business side, it disrupts the whole plan. And, for the customers, why would they give you another shot?
Hard decisions are irreversible and they have a significant impact on customers and therefore on business.
Framework for making hard decisions
This framework is made of 2 questions and 2 axes. First, we recognize whether the decision itself is reversible or not.

First, we evaluate whether the decision can be reversed or not without a lot of repercussions. If it is reversible, then we proceed with what seems good and revisit if required. It is like an iterative model of the agile world.
And when a decision cannot be reversed, we see where does the decision options fit in business benefit and customer severity graph.

Now, we try to place the options in one of the quadrants aligned to the axis, based on the information we have.
For the business benefit, it is relatively easy to figure impact on the long and short term. For example – A short burst on usage metrics or long term metrics like LTV or retention.
For the customer severity, it is slightly difficult to measure the impact customers would have because of the decision. Past data experiences provide a quick guide. Total number of customers impacted could be easy to use metrics to measure. A small discussion with Sales and Customer Management teams always helps.
It is important to understand that we are measuring the size of impact by severity, and not the positive or negative direction of impact. A great benefit to customers or a gargantuan negative experience, both for a small set of customers have an impact and falls on the same side of the axis.
I wish I could add a time axis here. With time, we can have more research and more information, however time is always limiting and will not drastically change the framework.
Now, quadrants:
1st Quadrant – High customer impact and long term business benefit
Options in this quadrant represent a sound choice. Taking decisions between the options in this quadrant would depend on several more factors, like business maturity, market position, and the organization’s mission. Amazon PMs would choose options with more long-term business benefits than Tata Cliq. However, in any case, preparing customers for the change in this quadrant is important here.
Since the impact is high, this decision should be communicated to the leadership, and because there is a long term business benefit, getting buy-in from them shouldn’t be a problem.
2nd Quadrant – High customer impact and short term business benefit
One should have a strong reason for a big trade-off that we are willing to make.
Documenting the whole case becomes much more important here. We should clearly state the outcomes and trade-offs, so in the future, we have strong evidence for the decision.
I would also recommend preparing for the worst-case here. If the short term benefits do not realize, then these highly impacted customers need to be pacified.
Lastly, we should start small and test the waters if the option of this quadrant is chosen.
3rd Quadrant – Low customer impact and short term business benefit
Options in the quadrant should be avoided. Decisions with these characteristics mean that the product strategy has not been discussed and well thought.
4th Quadrant – Low customer impact and long term business benefit
No brainer. If less number of customers care being impacted, and business is going to benefit in the future, then this option should be preferred.
Own the decision velocity – be aware on how long you have been dragging and what are the tradeoffs
That’s all! Product managers own the decision velocity. It is important to realize how long the decision have been dragging. In most cases, if the decision is reversible, understand the opportunity cost and move forward. If it is irreversible, hopefully this framework will help.
* In this particular case, Corona Virus brought in a quick regulation change that we were not prepared. We did not do customer research on our part as we planned to give more than what customers would have expected by designing a prorated logic that helped in calculations on the customer’s side. While we were stoked with our user-friendly solution (which competitors might not even provide), the complexities of prorating calculations dragged us down after a few days when new use cases emerged. Therefore, just after 2 days of sending our update, we had to send another update to nullify the previous prorated logic. Thankfully, our operations team handled beautifully by getting buy-in from all customers.
Other experiences I had – Cross-promotion of products for which I had a major disagreement. I believe promotional banners inside the paid product would not be helpful for the users. The customer impact as high as it was a change on their face, distracting them from their intended journey. The objective was to cover the deficit MRR. We tested it with a small set of customers and found the tactic ineffective.