Powering Product Management with Data-Driven Decisions

    The BizOps Manifesto has been released at bizopsmanifesto.org. One of the key values featured is this: Use data to make decisions, rather than opinions, judgment, and persuasion. In a previous post, we talked about the advantages of using data to manage your business (see “Data-Driven Product Management”). Now let’s look at how we can employ data-driven decisions to power organizational change.

    Establish Organization-Wide Alignment, Agility

    By harnessing BizOps, teams can establish cross-domain and cross-organization visibility. Through this approach, teams can begin to move beyond agile development and become agile businesses. This need for enterprise-wide agility is essential and intensifying. I’ve often described this concept via the analogy of coffee being poured into a cup. A lot of energy can be expended heating coffee, which will then be composed of fast-moving molecules. However, once the coffee is poured into a cup and exposed to cooler air, the molecules will immediately begin to slow and cool.

    To me, this is a great analogy of having an agile development team, while the rest of the business operates in a legacy, non-agile fashion. Agile teams will encounter guardrails and processes that stifle and slow their progress. As I stated in my blog post on “Lean Portfolio Management,” your teams will only stay as agile and fast-paced as the rest of the company allows them to be.

    Embrace Customer Centricity

    I often hear executives talk about how their organizations are pursuing ways to become customer centric. However, quite often, these efforts are viewed as one-time projects or initiatives. Today, that’s not enough. Customer centricity is at the core of any successful digital transformation, and it needs to be a foundational element, part of the organization’s DNA. (To learn more on becoming a customer-centric business, see my earlier post on agile business management.)

    Use Data to Reward and Inform, Not Punish

    Across the organization, it’s essential to use data to understand the reality of your business. However, it’s important to be careful about how the data is used. You want to build trust, not only in the accuracy and integrity of the data, but in the way people will use that data. Toward that end, focus on building systems for delivering intelligence, and using data for coaching and training.

    In my work, I’ve made it a standard policy to mandate that data will not be used punitively until data initiatives have been in place for two years or more. This gives teams enough time to trust that data won’t be used against them, which improves trust throughout the organization and gives you access to data that depicts reality—whether the news is good or not. Imagine if you had real data to make all business decisions, trade-offs, and strategic realignments? It truly changes how you can run your business.

    Use metrics to determine where people need further assistance, how to help setting a sustainable pace, and where impediments need to be removed. There is usually a lot behind a metric that is not trending how you want and I have never found the answer to be that people weren’t working hard enough.

    Align Outcomes to Build Trust and Remove Silos

    Historically, across the organization, individuals and teams have been given bonuses and other performance incentives based on distinct goals. For example, in any organization, it can be common to see IT staff being given bonuses for quality or SLA compliance, development teams tracking incentives on speed, and business teams being evaluated based on net-promoter scores or on-time delivery.

    It’s important to establish visibility that spans these different disciplines. With that in place, it’s then possible to track and ultimately establish unified performance incentives. Ultimately, the goal should be to establish the same metrics and objectives for everyone, including technical and business teams. Rather than pitting teams against one another, this fosters improved collaboration, trust, communication, and coordination. If one team fails, all will fail. All succeed together.

    Even if you can’t change the ranking or HR system used to evaluate people, you can change the factors that go into decisions. How people are incented often impacts their behavior and their willingness to show you what is really happening. We literally have decades of training in obfuscating reality from leaders: Phase-two SDLC plans that promise we will hit time, budget, and scope targets; status meetings where we are asked if we are done, without “done” being defined; status reports that show a status of green until a month prior to release; and so on. All of these things have been acceptable and in place for the past 30 years.

    Using a small set of dashboards for tracking both business and technology work and showing activities at each level in the organization, is a great way to power data-driven decisions and start to align technology teams with the business.

    Empower Staff to Focus

    In our day-to-day jobs, staying focused can be difficult and this is especially true now that we’re working from home. The more directions we’re pulled in, the less progress we make towards the objectives that really matter. That’s why it’s so important, particularly for strategic roles, to promote focus, enabling staff to dedicate their time and energy to specific projects and efforts. However, too often, individuals have roles that leave them handling multiple projects simultaneously.

    Think about it from a sports standpoint. Take the best athletes in a given sport, then imagine if they had to be playing for two or three teams simultaneously. They’d be less comfortable with any one team’s plays and schemes. They’d be slower and less productive. Ultimately, they’d be completely worn out.

    However, in the corporate world, that’s a scenario you see play out routinely. For example, I remember working with a team that had their three top architects spread across 12 projects. They were running ragged, working long days and weekends. They were burnt out, and the teams they supported were frustrated.

    The company didn’t have the headcount or budget to establish a 1:1 architect-to-project ratio, so what’s the solution? We established a schedule in which each architect would be dedicated to a single project over the course of a two-week sprint. An architect would be focused solely on project A for one two-week sprint, then focus on project B the following two weeks, and so on. This allowed teams to know when they would get the architects’ time, and that the time with them would be dedicated. The result: The architects’ productivity increased by 10 times and morale improved significantly, both among the architects and the project teams. Further, all these benefits were realized without an increase in staffing.

    Another way to increase focus is to do away with status meetings, and instead rely on dashboards and data to understand the status of a project. If the focus shifts to using the data that is naturally created through good agile practices, and you have standards and guardrails in place for teams (definitions of done, release criteria, and so on), you can dramatically reduce the amount of time both business and technology teams spend in meetings.

    In order to reduce distractions, really focus on what we call “the work about the work” as opposed to the work itself. Pay special attention to the amount of time, and number of times, your teams are disrupted by oversight meetings, emails for status, requests to fill out spreadsheets for status, and the like. The more time your teams spend writing or talking about what they are doing, the less time they are actually doing work.

    Escaping the Manual Status Reporting Trap

    If you still rely on manual status reports, long status meetings, and ambiguous definitions for completed work, instituting data-driven decisions and approaches can have a major impact on your business. Start by replacing status reports and meetings with data from your internal systems. Our ultimate goal is to build a pipeline of data from the top tier-strategy to the customer and back again, tracking every step in between.

    The first step is to make sure you have your strategic plans strongly tied to how the work is broken down. 78% of companies have said that the disconnect between IT and business units results in significant costs. First, make sure you aren’t creating long-term strategies and then expecting teams to magically break a nine-month plan into two-week sprints. You need steps in between, and guidelines in place, to ensure technical debt isn’t built up along the way. See a previous blog, “Establishing a Definition of Done,” for details.

    Image 1

    As an example, in the picture above, let’s assume the strategy will take 15 months to achieve. Our initiatives might be nine and six months each. Features should fit into a quarter and be named so that they’d be recognizable by both internal staff and customers. Then fit stories into two-week sprints. I used to give guidelines that, with testing, stories should last no more than three or four business days. This gives you breakdowns that are manageable and the thought that goes into breaking down the strategies often helps to validate that you have the right strategic plan in place.

    As your teams start to complete the stories, and the business accepts them as complete (since you also have a standard definition of done), you can watch the strategy come to life. Completed stories roll up to completed features, features to initiatives, and initiatives to strategies. And every level of the organization involved in the work should have the data they need to run their business.

    Too Many Strategies

    Another great place to start fostering data-driven decisions is by using data to validate that you aren’t investing in more work than can be done by your teams. I once had a company executive tell me that she had 35,000 developers and approximately 70,000 hours of time available each year. But she had no idea if she was giving the teams 50,000 hours of work or 150,000 hours of work when she funded their initiatives for the year. The initial estimates seemed to be accurate enough, but they didn’t know how long the work actually took.

    To combat this, some organizations are changing how they fund work. Instead of using dollars, they are using team capacity. When their capacity is full, they’ve spent all their budget. This takes more effort because it requires teams to do actual estimating for it to work. However, this approach can significantly improve estimates.

    Other organizations are looking at charts that track how much work teams are able to produce within a specific period of time and then tracking the ask vs. the output. These cumulative flow diagrams (CFD) tend to look like this:

    Image 2

    You can see that the volume of asks is rising at a much faster rate than the teams can complete work. Also, while teams are completing more (likely due to automation or improved productivity), the ask is far greater than the completion rate. At this rate, there is enough work in the pipeline to keep the teams busy for more than three years—and that is if nothing else is added over that entire time.

    When we first dig into this with companies, we find, on average, approximately 250% more asks than the teams can complete. Using a chart like this leads to conversations about customer value and prioritization, which we’ll leave for a future blog post, but it helps you see the reality of your business. Maybe your teams are moving as fast as they can, but the ask is far outpacing the work. This creates a situation where it appears “nothing is getting done,” even though teams are consistent and even improving their output.

    Data-Driven Decisions in Action

    When everyone is working from a standard set of data, the business impact can be significant. For example, I worked on the transformation of a large financial services business. At any given time, the business had around 30 global strategies in play. Ultimately, we were able to establish a management platform that powered data-driven decisions, and spanned all strategies across the entire company. We were able to deliver tailored dashboards at all levels of the organization. From developers to the CTO, everyone could track trends, view current status, and get actionable insights—all based on real, current, and accurate data. This had a dramatic, game-changing impact on how we ran the organization, and it allowed us to beat competition to market as well as make our customers happy.

    Conclusion

    Time and time again, I’ve seen how transformative it can be when teams across the organization start to fully leverage the power of data. To capitalize on this opportunity, strive to make sure you use data to foster customer centricity, inform and reward staff, align efforts with outcomes, and boost focus on the strategies that matter. When teams move from opinions and hunches to data-driven decisions, they can be more customer centric, focused, efficient, and effective. Ultimately, these data-driven teams achieve better outcomes and power enhanced business success.