Not enough effort has been made to understand the people using data products, what they want to get done, and the broader context in which they operate. I believe there is one emerging role that could radically reduce the gap between the insights offered by data analytics and the customers who could benefit from them. That role is the Data Product Manager.
Across the globe, businesses are investing billions of dollars in data analytics – with minimal return. Leading research and advisory firm Gartner has predicted that through 2022, only 20% of analytic insights will deliver business outcomes. Clearly, analytics initiatives are falling far short of the high expectations that many businesses have of them.
There’s a growing consensus that data organisations can have a bigger impact by shifting from service organisations to product organisations. Products are much more scalable and therefore can create greater impact. In their article, Emilie Schario and Taylor A. Murphy advocate running data teams like product teams, viewing their colleagues as customers using their data products. They define a “data product” as the collection of every piece of data, and the tools used to generate, access, and analyse that data within an organisation. I will adopt the same definition for this article. Additionally, in this article for the Harvard Business Review, the authors highlight how many businesses fail to become data-driven due to employee (or customer) challenges and recommend data teams approach this through a product development lens to improve business outcomes.
For this article, I will limit the scope to discussing internal-facing data products, and “customers” will refer to employees within an organisation. The practice of serving external customers is relatively more established and well served through the practice of traditional product management.
Leveraging product management in data analytics makes your team more impactful
Data organisations can run more effectively by considering key principles from product management. This shift in approach has the potential to benefit the wider organisation substantially and elevate the data organisation’s impact. These benefits include, but are not limited to:
- Better business impact, with a focus on business outcomes. Too often, companies aren’t able to define precisely what the data product is supposed to do for their business, and the end customers. Generating data products can create the illusion of progress, but without careful attention to outcomes, organisations miss an opportunity to meaningfully improve business results. Product management principles help to develop a clear purpose for data products, by focusing on what success looks like for customers, and how accessible intelligence could help them achieve their goals. This clarity has the added benefit of enabling effective communication with leadership to win their support for data initiatives.
- A deep understanding of customer problems, to better address their needs. Understanding what customers want to get done, and how they might go about that, will lead to solutions that fit their needs. For example, an analytical operational tool for the marketing department is more likely to be adopted by customers if it’s extremely user-friendly and customised appropriately to complement their capabilities and workflows. Without completing this last mile effort, analytics investments can go to waste.
- Clear priorities, to solve the right problems at the right time. Without a clear framework for prioritisation that aligns with near-term strategic objectives, data teams risk creating data products that aren’t relevant to the business at that time. Such products are less likely to be adopted and have the potential to waste significant human and financial resources. Getting priorities clear is often a messy process, requiring strong relations with stakeholders to negotiate constraints.
A natural question to ask is, why should we have a dedicated role for applying product management principles? Couldn’t other roles in the data organisation simply adopt them to reap the benefits? There is some merit in this suggestion. It would certainly be better than doing nothing.
However, having a dedicated role provides the capacity to really focus on customer problems. While this may sound simple or intuitive, this actually entails a lot of work, including developing a deep understanding of the problem space, building alignment with partners, synthesising information, strategic prioritisation, and a lot of communication. Taking lessons from traditional product management, we see a lot of value in separating the process of building the right thing (product) versus building the thing right (engineering). Currently, data teams rely on individuals to do both, but it’s rare to find such individuals that excel in both. Finally, a dedicated role means accountability for solving customer problems, and with that comes a focus on outcomes for the organisation, which are conducive to success.
The need for a dedicated role will depend on the maturity of the business in question. There are no hard and fast rules. But an important signal is a lack of outcomes from work produced by the data organisation and a sense that there’s potential to make an even greater impact on the business.
Truly understanding and solving customer needs drives usage and improves outcomes
In a previous role at a marketplace company, my colleagues (or customers) from the operations team were seeking help from the data team to make it easier for them to explore user data so they could more effectively match demand with supply through manual intervention, a common growth tactic in the early stages. Growth was a clear strategic priority for the business and on the data team, and we prioritised helping the business achieve this goal.
We took the time to understand their requirements and based on this, built a flexible dashboard tool that would provide them with the information they needed about users. A tutorial was provided on how to use the new dashboard with expectations that it would enable the team to self-serve. Several weeks in, the usage analytics showed staff weren’t using the dashboard as intended and were still asking for information that could easily be served through the dashboard. Through quick user interviews, we identified several UX issues and made changes to rectify these. But it didn’t lead to an uptake in usage. This was not a scalable situation for the data team to serve individual requests and, more importantly, it wasn’t enabling meaningful business impact. We went back with a carefully considered interview plan in search of a root cause. This led to the discovery that most of the operations team weren’t confident in using the tool themselves and needed more step-by-step guidance on-demand. The initial tutorial wasn’t enough for persisting the required knowledge.
To address this, we created walkthrough videos starting with top use cases for the dashboard and made them easily accessible. This allowed our customers to access these videos on-demand as they needed when using the tool. Over time, this led to increased usage rates and a decline in support requests. With the operations team empowered with the tools and resources they needed, we saw match rates (a key KPI for a marketplace business) more than double. This significantly impacted the business as it provided value for its users and fueled its growth engine as matches led to an even further increase in organic growth.
This story highlights the value of diving deep to understand the root cause of a customer problem – a core principle of product management. This deep understanding coupled with strategic priorities and a focus on outcomes leads to powerful business impact.
Data Product Managers bring specific internally-focused skills
Clearly, a dedicated Data Product Manager has the potential to create meaningful impact. But defining an emerging role or practice can be difficult. There’s value in calling out some key considerations when shaping this role.
As mentioned, the practice of managing external products, including external data products (e.g. recommendation engine for shoppers), is relatively more established through traditional product management. But managing internal products is still an emerging practice in organisations.
The core skill sets needed to manage these two types of products are largely the same, but there are distinct skills and strategies required to manage internal products, due to the nature of customers and their relationship with the product. The internally focused Data Product Manager should have these skills:
- Demonstrating business value. A persistent challenge for enabling teams (i.e. Data, DevOps) is demonstrating business value. They can be several layers removed from revenue generation, which can make it difficult to ask for more resourcing. Data Product Managers need to have strategic communication skills that allow them to communicate the value being created throughout the organisation.
- Change management. Introducing and integrating internal products require new behaviours and processes from customers compared to external products. For example, a go-to-market plan for an internal product will draw upon change management techniques, whilst the plan for an external product would utilise growth campaign techniques.
- Evangelism. To have the largest impact on the business, internal customers need to understand what products are available and the data team’s abilities to solve customer needs. Strong evangelism and comfort promoting the team’s work internally enable this.
- Complex stakeholder management. For internal products, the process of establishing roadmaps and prioritising would differ, as access to customers is much more attainable. This provides an opportunity to go much deeper into problem areas and understand customer priorities better. This may seem like it would make the process easier but due to the depth of information, proximity to customers, and competing priorities across different customer groups, it’s likely to add more complexity, requiring skilful cross-functional stakeholder management.
Other factors to consider are specialist knowledge requirements and the industry sector. Some roles may call for people with specialist knowledge in a particular area, such as Machine Learning or Business Intelligence. Also, the focus of the role will vary, depending on which sector the organisation operates in: for example, you would expect a greater emphasis on privacy issues in a healthcare company, while an e-commerce company would value quick feedback loops through experimentation.
There is much to be gained from applying best practices in product management within the domain of data analytics. Doing so by introducing the new role of Data Product Manager would harness complementary skill sets for the benefit of the data team, their colleagues, and external users. I firmly believe that adopting these values and approaches from product management will help run more effective data organisations. I would love to hear your thoughts: do the problems presented here resonate with you? How could your organisation benefit from a Data Product Manager? Please send me an email, or join our Slack channel to share.
