Why is prioritisation important?

As a product manager, you will always have constraints – less number of engineers, limited time before launch, a fixed amount of money/resources, etc.

In other words, you will have more things to do than possible with the resources available.
That is why you need to prioritise. You need to select the most important thing(s) to do given the constraints.

What does prioritisation mean?

Prioritisation: the process of doing the thing(s) that gets you closest to your goal, given the constraints.

How to prioritise effectively?

1. Keep it simple: frameworks sound good in theory but are tricky to apply in practice. Create frameworks and processes that work for you and your team. Stick to it.

2. Always know the goal: and use it goal as a guiding light. If there is no other information available, select the one item that will move you closest to the goal in the least amount of time. (See example below)

3. Use data: but don’t obsess over it.

Why is prioritisation tough?

I will use an example to make my point here

Let us assume that we’re a B2B Ad platform, and that our goal is to increase ad revenue.

Ideal world scenario:

1. Feature A needs four weeks to build and increases revenue by 15%.
2. Feature B needs four weeks to build and increases revenue by 5%.

In this case, prioritisation is straightforward. I would select A.

In the real world, nothing is straightforward.

Real-world scenario:

1. Feature A needs four weeks to build and increases Ad CTR (click-through rate) by 5%
2. Feature B needs four weeks to build and increases the sales team’s efficiency by 30%

In this case, the cost is comparable. But, the impact isn’t. This is because of two reasons:

1. None of the features have a direct revenue impact sizing.
2. The unit of impact in both the features is different – Ad CTR vs. team efficiency

In this case, this is what I would do:

Feature A:
1. Look at existing customers’ data. Then segment customers by the average CTR
2. Compare monthly revenue between customers with higher CTR and those with lower CTR.
3. Approximate the incremental revenue per % point of incremental CTR. Something along the lines of: “Customers with a 2% CTR have an average monthly revenue of $500, and customers with 2.2% CTR have an avg. monthly revenue of $550” In other words, every 10% of increment in CTR leads to a 10% increase in monthly revenue
4. Estimate the number of customers that Feature A will impact. Let us say that it is 500. From data, determine the average monthly revenue per customer (let us say that it is $200 a month)

Then calculate the incremental revenue = 500 customers * $200 * 5% = $5000 per month

For feature B:
1. Calculate the number of hours a salesperson works every month (say 150)
2. Determine the number of customers they acquire in a month (say 10)
3. Find (from existing data) the average revenue of customers acquired (say $200)
4. Find the total number of salespersons impacted by feature B (say 15)

Use the above to determine the revenue acquired per hour per salesperson = 10/150*$200 ~ $13 per hour

Feature B saves 30% time = 30%* 150 = 45 hours

Total incremental revenue for feature B = 45 hours *$13 * 15(salespeople) = $8775

So, I would prioritise feature B.

To summarise – the point of this example is not to teach you how to do the math, but the point is to note that:
1. Prioritisation does not have to be scientific or accurate to the last decimal
2. It does not have to follow a specific framework
3. As long as you know the goal, you can easily reach an accurate impact sizing
4. The more data you have, the better. But the absence of some data does not necessarily lead to bad prioritisation.