Data-driven prioritization framework

I integrated live data into the prioritization framework and so can you!!!

Photo by Stephen Phillips - Hostreviews.co.uk on Unsplash

Prioritization is one of the regular chores of a product manager. When a product is growing, you will get tons of requests from customers and stakeholders on daily basis. As product managers, it’s your job to prioritize ruthlessly to ensure you and your team are working on the most important things.

I was working on a SaaS product, which was in a growth phase. My team was getting hundreds of feature requests and bugs from customers, the customer success team, the sales team, account managers and other teams.

The manual triaging process was tedious for me and inefficient for the team. Therefore, I knew that I had to identify a process of managing these requests automatically.

What was I looking for?

A process or framework than can incorporate

1) Business and product direction: To not only prioritize the highest impact for customers but also ensure it aligns with business and product direction.

2) Data-driven: Identify a framework in which factors are quantifiable which would be vital in getting buy-in from all the stakeholders.

3) Lean: Identify a process which is lean. Like when a new feature is added, you don’t need to manually adjust or restart the priotization process over again.

Current prioritization frameworks and why they were not solving my problem

Note: I was looking for frameworks to prioritize features, pain points and bugs for an existing product (not for an MVP or new product launch).

Let's take a look at some of the popular priotization frameworks and why they were not enough to solve my problem

1. Impact vs Effort
   This framework was prioritizing features based on high impact and the tasks which would take the lowest amount of effort to implement.
   Pros:
   · Effort is quantifiable
   Cons:
   · What’s considered a high impact could be subjective and qualitative.
   · Each new task will need manual effort to assign impact and effort
2. Value vs. Complexity Quadrant
   This framework was prioritizing features based on value and implementation complexity.
   Pros
   · Value takes both user and business needs
   · Complexity is quantifiable
   Cons:
   · What’s considered a high value could be subjective and qualitative
   · Each new task will need manual effort to assign value and complexity
3. Weighted Scoring Prioritization
   It uses numerical scoring to rank strategic initiatives against benefit and cost categories. (source)
   Pros:
   · All the factors are quantifiable
   · Factors can be aligned with business objectives
   Cons:
   · Each new task will need manual effort to assign scoring
4. RICE
   RICE allows working on the initiatives that are most likely to impact any given goal. This scoring system measures each feature or initiative against four factors: reach, impact, confidence and effort. (Source)
   Pros:
   · All the factors are quantifiable as you can assign scores to them
   Cons:
   · Manually assigning scores to Impact can be subjective
   · Manually assigning scores to Confidence can be subjective

So how did I solve the problem?

1. I used a combination of RICE (Reach, impact, confidence, effort) and Weighted Scoring Prioritization.
2. Along with integrating real-time data
3. Assigned weights according to business objectives and product-prioritized persona.

The framework I created was
Reach[W*1(number of customers using feature A] X Impact [W*2(impact on individual customer + impact on other customers) + (impact on business)] X Confidence[W*3(Company size)]

W=weight
Example for real-life scenario:

These were all quantifiable factors, integrating real-time data from the relative data sources.

Customer name and feature name were made mandatory while reporting a ticket.

When a ticket was created, the priority would be assigned automatically because all the logic and total score (from 50) were calculated behind the scene.
Priority field logic:
More than 40 points assign priority > High
More than 25 points assign priority > Medium
More than 10 points assign priority > Low
Less than 10 points assign priority > Very low

Hence, I was successfully able to

✅ Not only prioritize the highest impact for customers but also ensure business and product direction.
✅ Implement a framework in which factors are quantifiable.
✅ Implement a process which is lean. Like when a new feature is added, I don’t need to manually adjust and restart the priotization process over again.

💡The good news is you can convert existing priotization frameworks to data-driven too.

Below are some of the data points that can help you integrate real-time data in the priotization frameworks

Tip: You can add the following data points in Value, Impact, Reach, and Confidence or while applying weight to certain factors

1. The number of customers currently using that specific feature.
   For example
   If 90% of customers are using Feature A, then assign 5 points.
   If 10% of customers are using Feature A, then assign 0.5 points.
   Data Point:
   Add the customer name and Feature name field in the ticket
2. For a specific platform
   · Mobile VS desktop VS iPad
   · Mac vs windows
   · Chrome VS safari VS firefox.
   For example
   If you are a mobile-first company then assign
   1 for mobile
   0 for desktop
   Data Point:
   Add the field for the Platform in the ticket.
3. Reported for a specific page: Number of daily active users on the page.
   Data Point:
   Each product has a different name for its different interaction levels. “Add interacted at/Page name“ field in the ticket
4. Type of persona prioritized for business strategy
   · Enterprise VS medium VS small
   · Merchants VS consumers
   · Types of preferred partners
   For example
   If your business wants to increase the number of enterprise customers then a pain point reported by
   Enterprise client= 3
   Medium-sized business=2
   Small-sized business =1
   Data Point:
   Add the customer name field to the ticket. Calculate at the back whether this customer belongs to an enterprise company or not and assign points accordingly.
5. Revenue generated.
   · Revenue generated by a specific customer.
   · Revenue generated by a specific partner
   For example
   A pain point reported by the top 5% of highly profitable customers will have a high weightage or points as compared to a customer who is contributing less to the business in terms of revenue.
   Data Point:
   Add the customer name field to the ticket. Calculate at the back with connection to the data source (e.g salesforce) how much this customer or her company, is paying us.
6. For specific business metrics
   · Acquisition
   · Activation
   · Retention
   Is the task or issue raised by a customer who is new, existing or who is in the sales funnel?
   For example
   When your company wants to increase expansion or retention, then issues reported by an existing customer will take more weight as compared to the one in the funnel or in the activation stage.
   Data Point:
   Add the customer name field to the ticket. Calculate at the back with connection to the data source (e.g salesforce) that is the customer still in the sales funnel or is our existing customer.
7. How much issue or task is critical for a specific customer? At what step customer is stuck?
   This is especially for SaaS companies whose services are being utilized by others.
   For example
   If a customer cannot complete a further 80% of a task, its priority will be high as compared to the customers which were not able to complete 5% of the tasks.
   Data Point:
   Add the feature name or API name field to the ticket.
8. Feelings for customers due to the presence or absence of specific feature [glad, mad, sad]
   This could only be achieved if you have done an NPS survey or asked for a review on various journeys or for features.
   Data Point:
   Add the feature name or API name field to the ticket
9. Business-critical or show-stopper for business.
   Is it a regulatory or compliance issue?
   Assign all the points in your bucket and also generate a slack or email notification to alert the respective team.
   Data Point:
   Add a field “Is regulatory or compliance”, with Yes or No drop-down or radio buttons.
10. Recommendation: Only apply the Effort factor when data-driven prioritization has been done.

Because if its a compliance issue or is reported by your top customer who is contributing to 15% of your revenue or 95% of the users are being affected by that pain point, no matter how much effort is needed we must prioritize that task.

I acknowledge that there will be times when you will have to decide based on a hunch or past experiences. Especially when you don’t have any data, no key persona or metrics prioritized on a high level. Also, there will be some edge cases which you would need to cater on a case-by-case basis.

But when you are in a situation where you have to prioritize each and every day, then manually applying prioritization does not remain scalable.

For any suggestions, feedback or queries, please feel free to comment or contact me at my email nidasaleem333@gmail.com.