March 26, 2021

Knime CEO Michael Berthold on How To Get More Yes at Executive Meetings with BI

Written by

For many, the shift to a data-first company plan can be terrifying. Some are afraid of what the data will find and others fear their process of making improvements on gut feeling and hindsight will be discovered. But the promise of a data-driven company is that the numbers back up your decisions when you set goals according to the data.

To help you get ahead in a data-driven company, you’ve got to centralize your departmental plans around the company goals. We spoke with Michael Berthold, the CEO and co-founder of KNIME, an open source data analytics company, to understand how managers from all departments can put data to work in ways that executives will appreciate.

Use the company’s goals and work backwards

A data-driven company will spend a lot of time defining the goals and metrics that they use to measure success.

“Mapping goals to numbers/metrics is most important here; if the reports showing progress toward company goals miss the underlying intention, the entire organization will start optimizing for the wrong goal. It is critical to always understand the mechanics behind all numbers to not only make sure the number is hit but also that the actual goal behind the number is achieved. One example: Measuring the success of a development team is almost impossible without constantly looking behind the numbers,” says Berthold.

How to work backwards from the company goals

A manager’s job is to align their team’s priorities with the company goals and metrics and execute on the tasks that will drive progress toward those goals. If the company goal is to make 10% more revenue, you may ask where your team’s locus of control over building revenue is most pronounced? What metrics can we improve to achieve those goals?”

When you begin with the end result in mind, you can improve the data collection. If the revenue numbers need to go up by $1M this quarter,

  • How many sales will it take to produce that change?
  • How many more meetings does the sales team need to have to produce those sales?
  • How many more leads does marketing need to hand over to sales?
  • In order to generate that many leads, how many new assets does the marketing team need to prepare, and how much design time will be required?

Each of the teams can break their portion of the goal into defined parts.

  • Design will need to produce 8 new display ads and design 4 new white papers.
  • Marketing needs to write 4 new white papers and enough email campaigns to drive 500 new leads.
  • Sales will prioritize the leads and begin outreach campaigns that produce the required meetings.

When you get really specific about the project and can quantify the outcomes that you expect, you can then put those metrics on a dashboard that the entire company can view to track progress on the goals.

Build forecasts with predictive analytics

“It’s important that managers understand the data they use to create a forecast and the dependencies in the data. Always question forecasts that look suspiciously positive — they tend to be wrong,” says Berthold.

This comes down to having a good understanding of your business and the market factors that will make the most impact. He advises companies to keep it simple: “Managers should not rely on overly sophisticated ML/AI methods — forecasting is usually a very basic process of identifying a few key drivers and how they affect the desired outcome. Forecasting requires business understanding; it’s not purely data driven predictive analytics.”

How to build forecasts

Building a forecast is a tricky thing. You don’t want to overshoot the mark with numbers that are unattainable, but you also don’t want to be accused of sandbagging to make the team look good. How do you walk that line appropriately? With predictive analytics.

If you’re building your Q3 sales forecast, write down your known variables like previous Q3 buying trends, current economic conditions in your industry, and your own staffing levels. Then consider how unknown variables might affect these. Above all, keep your forecasts within reasonable limits.

Work across departmental lines

This shouldn’t be a problem, but somehow it always is. Remind your team members that the company’s goals are shared goals. If two departments can work together closely to achieve the goal, then they should.

Berthold says you should “Promise something in return: We can help make your team understand better what works/what doesn’t work. We can help you assign resources more productively. And make it clear you don’t want to use the data solely to ‘measure performance’ but as a tool to provide insights to improve.”

How to do interdepartmental data work

When planning for your project, make sure to get the right people in the room from the beginning. This may mean starting with an interdepartmental team with a data analyst, sales, marketing, design, and development all in the same place to help outline the project’s goals.

Then give credit where credit is due. Build reports that show the work that both of your departments has contributed to the shared goal. Show how the combined work is moving the initiatives forward, raising the revenue, or reducing customer churn. Make a point of naming all of the contributors, no matter how small their part in the work.

Consider adding names and leaderboards to company dashboards. Which email marketer has the best open rate? Which salesperson has held the most meetings? These types of dashboards combine key tools that improve morale and drive results: accountability, recognition, data-driven decisions, and transparency into how the company goals are achieved.

Present a united front

If your executive team has outlined quarterly or yearly priorities for the company, they have shown you what they want you to pay attention to. Your team and the teams you collaborate with now need to show them reports that speak directly to those priorities.

Berthold recommends doing a brief retrospective to provide context. “I would always show past trends to ground the discussion in reality in addition to showing progress toward the bigger goals. And right next to it, I would add explanations for both scenarios: surprisingly positive progress as well as the not-so-good numbers. Always explain what numbers mean so you don’t get pulled into discussions of the type ‘why is that 7, not an 8.’”

How to present a united front with data

Your reporting should present the most important numbers front and center, no matter the contributor’s department. How much revenue did you bring in this month, and what ratio of the goal did you achieve? If you’re presenting a multi-page report or several slides, use later slides to break out the details. Executives want to see how the team has performed toward a goal, but also what they did to achieve the work. Answer the questions you know the executives will ask:

  • What decisions were made this time that made an impact?
  • How can you repeat this next quarter?
  • When can we expect to see results?

Good reporting improves with good visualizations

You can report well with spreadsheets, but a true business intelligence platform will speed up your reporting processes and help you build valid forecasts that convince executives. All of this comes with the added benefits of better organized data, faster business insights, and reporting at your fingertips.

Find the best business intelligence software for your company’s goals by using our Product Selection Tool. Enter your requirements and one of our unbiased Technology Advisors will provide you with a short list of BI vendors that meet your needs.


Knime CEO Michael Berthold

Michael Berthold is CEO and co-founder at KNIME, an open source data analytics company. He has more than 25 years of experience in data science, working in academia, most recently as a full professor at Konstanz University (Germany) and previously at University of California, Berkeley and Carnegie Mellon, and in industry at Intel’s Neural Network Group, Utopy, and Tripos. Michael has published extensively on data analytics, machine learning, and artificial intelligence. Follow Michael on Twitter, LinkedIn, and the KNIME blog.