In the previous chapter, we looked at how data-driven product management adds value by helping product managers measure whether business targets are being met and if customers are engaging with their product.
In this chapter, we will look at how to apply a data-driven product management approach in practice, by analyzing the four main steps that comprise this approach.
The data-driven product management cycle
Let's look at 4 steps for data-driven product management:
Decompose - Break down the organization's strategy into measurable outcomes.
Divide - For each product team in the organization, decide the targets of their respective product in terms of contribution to these measurable outcomes. If you add up the results of all the product teams in the organization, the result should be what the organization as a whole needs to achieve.
Determine - Build the product/service by determining and prioritizing the features that achieve these outcomes.
Discern - Measure product performance to understand if these outcomes are being achieved.
1. Decompose strategy into measurable outcomes
Strategy is a plan of action designed to achieve a long-term or overall aim. Professor Paula Jarzabkowski, author of Strategy as Practice states that:
Strategy generally involves setting goals, determining actions to achieve the goals, and mobilizing resources to execute the action.
The first step of data-driven product management is to take a strategy and break it up into measurable outcomes.
If your company's strategy is "to be the global market leader in online payments in the next three years by introducing our existing products into new markets," then how could you break up such a statement into more measurable outcomes?
Crafting a set of measurable goals for each required objective would be a good first step. For example, to be the global market leader, you would anticipate that the following would have to happen in the next 12 months:
Our revenue must exceed $100 million a year.
Our average-revenue for current clients must increase by 30%.
Our fraud levels should be under .01% of transactions.
We must retain 98% of our talented staff each year.
Objectives and key results
The acronym OKRs stands for objectives and key results. OKRs are a management approach to measuring whether the long-term goals of the organization are on track to being met by breaking them up into short-term objectives (sub-goals) and key results (clear milestones for each objective to measure progress).
An objective answers the question, “Where do we want to go?”
A key result answers the question, “How will we know we're getting there?”
As an example, let’s consider customer satisfaction as a priority.
Your objective could be that 95% of customers are happy.
Key results could be that:
Less than 1% of monthly customers cancel.
Fewer than 1000 customer support calls are received per month.
Customer satisfaction surveys indicate over 95% happiness.
These key results are then measured over time to see if your strategy is successful.
You can see an example in the diagram below.
SMART goals
When setting target outcomes or goals for your business, it is important that they are measurable.
For this reason, effective businesses set SMART goals:
Specific - A goal should not be vague or difficult to understand. Signing up ten new customers is very specific and easy to understand.
Measurable - A goal like "improving the customer experience" is vague and hard to measure. A more measurable goal would be to "reduce the wait times on our support phone line," for example.
Attainable - Who doesn't want to make billions of revenue every year and be the market leader? Is it realistic for your five-person team? Goals should feel like they can be accomplished.
Relevant - It's important that defined goals impact the core considerations of the business. A goal like "grow our Twitter followers to 1 million people" might be strategically important. If it is, you should be able to clearly articulate why.
Time-Bound - Achievable goals require deadlines. If your plan is to"sign 10 new clients this year," then you will know if your goal was met on December 31st.
2. Divide the outcomes among product teams
Because many organizations consist of many products and many product teams, it is important that each team is given targeted outcomes for each particular product.
Example
Let's take an example where the organizational strategy for the next year is to "acquire 100k additional paying customers." Assume that the organization has three product teams.
The responsibility for achieving this organizational goal could be split up as follows among the product teams:
Product team 1 will aim to acquire 10k additional paying customers.
Product team 2 will aim to acquire 60k additional paying customers.
Product team 3 will aim to acquire 30k additional paying customers.
Result
Each product team now has its own set of targets!
3. Determine product features
Now that the products have target outcomes, the product manager can decide which features will help them achieve their goals.
If a given product team needs to acquire 10k new paying customers, then questions like the following are suitable to ask about each potential feature before the product managers decide to make it a priority or not:
Will this feature help us attract new customers?
Will this feature help us attract new paying customers?
Will this feature help non-paying existing customers decide to make a purchase?
By having a measurable outcome to achieve, this product team is already measuring the work they do against whether it helps them achieve this outcome.
4. Discern if the product is performing/achieving desired outcomes
Once the product manager has decided on the features that will help his team achieve its goals, it's important to build and measure these features to see if they impact customer behavior.
If you built a new referral scheme that let users get some free months of premium membership (if their friends signed up and paid), then you would want to look at:
Were users recommending their friends?
Were friends then visiting our product or website as a result?
Did those friends create an account? Make a payment or purchase?
How many friends were being invited per user?
In the diagram below, you can see that this allows you to get data when you release this feature. You can examine this data to gain insight into whether the feature was successful and whether customers behaved as you expected them to.
This analysis is then used to improve the feature. Or, can be a factor in how you tackle other features in the future.
In the next chapter, we will look at how to choose a set of metrics that represent "success" and how monitoring them can tell you when a feature has made the contribution that you expect. We'll also look at how you can react when users do not behave as you had expected!
To get a full view of what the data-driven decision-making process for three product teams would look like, you could visualize it as being something like the diagram below!
Summary
Data-driven product management adds value by helping product managers measure whether business targets are being met and whether customers are engaging with their product.
The process can be explained by the "Four D's" of data-driven product management:
Decompose the organization's strategy into measurable outcomes.
Divide measurable outcomes into specific outcomes for each product team in the organization.
Determine features that each product team can build to achieve their team's target outcomes.
Discern if these outcomes are being achieved by analyzing data and using this insight to improve features.
Additional resources
Learn more about OKRs on this blog post on Medium.
