Why Ad Tech Companies Need a Data-Driven Product Strategy

Opinion: It’s guaranteed that a data-driven priority list will not mirror any one stakeholder’s top 10

Most of the companies I’ve worked for in advertising and media technology take a top-down approach when deciding what products to green-light. A top executive will come back from a conference jazzed about the latest buzz and call an emergency team meeting. “I know we still need to finish ABC, but we need to jump on XYZ immediately! How soon can we get it done?”

The exec will explain how XYZ is required to “take us to the next level.” They’ll say that if so-and-so releases it first, we’ll be out of business. Any early comment about how resources are already stretched will be met with shouts of, “We’ll hire more people,” or, “We’ll just clone Susan, ha ha ha!” When presented with strong top-down product direction, stakeholders tend to fall in line and rally around the ideas presented by leadership. We all want to please the boss.

Even when a more measured, less-reactive product review is conducted, shiny new things get disproportionate time and attention. Less-glamorous ideas are given short shrift or ignored altogether. Enhancements to existing systems take a back seat to new efforts.

While a top-down product directive is better than a total lack of vision, it can lead to wasted resources and team whiplash. It’s also possible that innovation is eating into potential revenue and vanity projects are negatively affecting your bottom line.

Why not put the hottest ideas to the test and see who’s right about top priorities? If your organization’s leadership is plugged into the company and industry, many of their top-down ideas will bubble up. But it’s guaranteed that a data-driven priority list will not mirror any one stakeholder’s top 10.

By following the data-driven process outlined here, your team will be fully invested in the outcomes and excited to be part of a shared vision. Here’s how it works:

Stakeholders—the ‘who’

First you must identify a group of leaders who can speak for key constituencies within your organization. This often involves the heads of sales, business development, tech, client services, partnerships, marketing, finance, legal, etc. These folks comprise your product strategy team and should be fully engaged in the prioritization process. They must actively solicit ideas from their respective groups and come into the process prepared to articulate a strong point of view.

Ideas—the ‘what’

Bring your product strategy team together for a conference. Lock everyone in a room for as long as it takes, and don’t let them out until there is a complete list of every possible product feature or idea.

At this stage, there’s no debating priority or pitting one idea against another. There are no bad ideas, and the goal is to get everything out in the open.

It sounds crazy at first, but most of us come to the table with lists of projects already in motion, a backlog of ideas on hold and some out-of-the-box concepts we’ve been mulling over.

The discussion can be facilitated by functional area, systems/subsystems, roles, organization structure—whatever works best for your group. The result is a firm list of ideas and a shared understanding of each goal and potential scope. Even after exploring the outliers—“Why don’t we shift gears and get into a completely different core business?”—each participant will eventually agree that all of their ideas have been captured.

Ranking factors—the ‘why’

Now you must choose four to eight ranking factors that speak to the needs of your organization. Consider what matters most when allocating resources this quarter. Time savings and efficiency? Revenue from new and existing clients? Competitive necessity? These will form the basis of your survey of stakeholders, with participants ranking each factor on a five-point scale.

This generates an empirical matrix of ranking factors that can be sliced and diced for robust priority analysis. You don’t need to decide now if efficiency is more important than revenue, since weights will be manipulated later when analyzing results. Take time to describe each ranking factor in detail and give real-world examples of what would score a one and what score a five on the scale. Get buy in from stakeholders on which ranking factors to use and what they mean

Once your ranking factors are set, you now have the lenses through which all ideas will be viewed and a sound framework for choosing priorities.

Figure 1: Training for a Ranking Factor

The survey—the ‘how’

When authoring your survey, take time to clean up the list of ideas and develop a concise description for each. Follow up with subject matter experts to review their respective ideas and cross-reference with the condensed descriptions to avoid losing something in translation.

When presenting the survey to participants, the first section is all about training. Start with a walkthrough of your ranking factors. After the training module, present each idea one at a time, with a matrix of ranking factors for input on a five-point scale. Collect responses using a web survey tool that offer easy exports of raw data, like SurveyMonkey.

Figure 2: Matrix of Ranking Factors

Here are some helpful survey-taking tips for participants tips to get your stakeholders focused:

  • Think about your job responsibilities when reading about each idea. How does this fit into your day-to-day routine? Why do you care about this?
  • Focus on your list of clients and stakeholders when ranking items one through five. How are the people most important to you affected?
  • If you have absolutely no experience with an idea or a ranking factor for an idea, you may mark your answer, “don’t know”. But please use this sparingly, and guessing is encouraged.
  • Answers will be combined across the entire organization for analysis, so don’t stress too much about any one answer.

Your first instinct is usually correct: Go with it.

The results—the ‘when’

To generate the list of top priorities, start by aggregating average rankings by all participants into a spreadsheet. Use weighted averages to derive a total score across all ranking factors and sort by the score. Voila: You now have a data-driven product strategy.

Figure 3: Analyzing Results

Share the list with the team, along with all of the source material used to reach conclusions. Let stakeholders manipulate the spreadsheet to test the model and gain confidence in the process. Allow individuals to compare their own ranked responses against the combined final list. How right was the CEO? How did that buzzworthy idea rank against everything else? It will be clear why some items fell lower in the list based on average scores and the final weighting of ranking factors.

Figure 4: Final Weighting of Ranking Factors

Rinse and repeat

The best part of a data-driven product plan is that new ideas can be scored and incorporated into overall rankings without needing to re-take existing surveys. New ideas take their appropriate place on the roadmap and don’t cut the line for the wrong reasons. The always-fresh product roadmap is available for execution teams to pull new work off the shelf as they free up.

Roy Firestone is vice president of product at native advertising provider MGID.

Image courtesy of from2015/iStock.