Talent analytics

How the IMPACT Framework Can Help You Solve Your Company’s Toughest Business Problems

Three in four talent professionals say that people analytics will be a major priority for their company over the next five years, according to data from the Global Talent Trends 2020 report. Despite this, many companies are not yet prepared. More than half (55%) of the respondents admit they still need help putting basic people analytics into practice.

Statistics from surveyed talent professionals in Global Talent Trends 2020:  55% of talent professionals say they still need help putting basic people analytics into practice.  73% say people analytics will be a major priority for their company over the next 5 years.

While many talent professionals feel comfortable with their ability to collect and maintain data, the next steps can present a challenge. Only 39% say their organization is good at actually analyzing the data — and even fewer feel confident acting and capitalizing on the insights they gain.

Statistics from a bar graph found in the Global Talent Trends 2020 report:  Title: How companies rate their own people analytics performance Subtitle: Percentage of talent pros who rate their own organization as good, fair, or poor at the following stages of people analytics mastery.  Collecting: Good - 44% Fair - 34% Poor - 22%  Maintaining: Good - 47% Fair - 33% Poor - 20%  Analyzing: Good - 39% Fair - 36% Poor - 26%  Acting: Good - 37% Fair - 37% Poor - 27%  Capitalizing: Good - 29% Fair - 35% Poor - 37%

To help your company master every stage of the process, LinkedIn’s Rebecca White, director of people analytics, and Katie Sittler, a senior people analytics associate, shared the framework their team uses to structure its workflow. By breaking each project down into six stages — Identify, Measure, Plan, Analyze, Communicate, Track (or IMPACT for short) — you can make people analytics feel more manageable and ensure you’re asking the right questions along the way.

Here’s how to put the IMPACT framework into practice and use it to gain actionable insights on issues like employee attrition. 

1. Identify: Make sure you understand the precise business issue you’re trying to solve

For Rebecca and Katie, the earliest steps in the framework are the ones they spend the most time on. These steps lay the foundations for a successful project, so even though the analysis stage might be the most exciting, it’s important not to rush in before you’re ready.

That’s especially true when it comes to identifying exactly what you’re trying to evaluate. If you can’t define the problem, you’ll have a much harder time trying to solve it. 

Say a manager asks you to look into attrition data. That’s a very broad request that was almost certainly triggered by a specific incident — like one department’s turnover being much higher than others. If you don’t know that, you might spend a lot of time digging into data that the manager simply doesn’t care about, because it isn’t related to the challenge they’re facing. 

“There are all these different layers to a simple request,” Katie says. “It's usually related to a direct key performance indicator (KPI) that we track that they're worried about.”

“Get clarity from them on what they’re seeing,” Rebecca advises. “What's prompting them to ask this question, and what do they think is happening?”

Once you have a clear idea of the business problem you’re trying to solve and the true scope of the request, you can come up with a hypothesis about what’s causing it that you then prove or disprove. The manager who made the request probably has an idea about the root cause, and through your analysis, you’ll be able to go back to them and tell them whether their instincts were correct or if other factors are at play. 

“Hypothesis creation is generally a co-creation with our partners,” Rebecca says, “so that they're able to point us in a direction to say, ‘Here's what I think might be going on. Can you dig in?’ That gets us a lot further down the road than if we tried to look at everything.”

As you’re coming up with your hypothesis, be sure to account for any data governance, privacy, or security issues that might arise. In some cases, these issues can stem from the nature of the requests themselves. If you’re not sure if a people analytics project would be compliant, always run it by your company’s legal team before you get stuck in. 

For example, Rebecca says her team has been asked questions like “Can you tell me the names of the last four people who left and why they left?” Sharing this kind of granular information, however, is not aligned with the high standards of trust LinkedIn has set around the privacy of our member and employee data.

So it’s your job to steer managers towards a higher-level problem-solving mindset, and that starts with educating them and having strategic conversations about the issues they’re trying to solve. 

2. Measure: Understand the strategic priority of the issue you’re trying to solve

Before diving into any analysis, LinkedIn’s people analytics team takes the time to understand the magnitude of each request and how much of a priority it really is. Without taking this step, your team can quickly find itself overwhelmed, and the most urgent requests can fall by the wayside. 

“The amount of things that people want to know about is always more than we have time to do,” Rebecca says. “So being vigilant about scoping and prioritizing is really important.”

Some requests can be highly time-consuming to complete, but may not deliver a proportional impact on the business. Others may not be particularly urgent. Managers might not think to mention this upfront or may not realize how busy your team is, so probing them for more information is crucial. 

––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––

Related: 9 Ways to Become a Strategic Talent Advisor to Your Business, According to John Vlastelica

––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––

“A lot of times, people are just curious,” Rebecca says. “That can take up an entire team's bandwidth. But by being really focused on making sure we're answering the most impactful, most important questions, for everything else, we're able to say ‘I can't do that for you because I'm focused on these other things.’ That helps people to understand.”

LinkedIn is also working to build more self-service infrastructure into its people analytics function. This will allow managers to satisfy their curiosity more easily by giving them easy access to certain metrics and reports. 

“My biggest piece of advice to any new people analytics team is to invest in your foundations and invest in reporting self-service,” Rebecca says. “That frees our time up to work on more strategic things.”

Putting this time back into your teams’ day can also make it easier to improve their job satisfaction and attract the talent you need — whether it’s seasoned analytics professionals or members of the talent acquisition team who are interested in working with data.  

3. Plan: Outline your plan of action and figure out what data you need 

When you have a clear hypothesis and know that it’s a business priority, the next step is to determine what data you need to look at and how you’re going to perform your analysis. 

For many business problems, such as evaluating the effectiveness of your recruiting channels, the data you need can likely be pulled from your applicant tracking system (ATS). Rebecca says that many people analytics teams ultimately get started this way, with one talent acquisition professional exploring their ATS’s built-in reporting functionality and getting curious. 

“Usually, some smart person who likes spreadsheets will start pulling data from the ATS and do some tinkering and analysis,” she says. “Once they've shown that ‘Oh, there's some really interesting insights we can pull from this, this is really helpful,’ then companies will start to get more serious about investing in a people analytics team.”

Your ATS isn’t the only place where you can source data. For a problem like attrition, LinkedIn’s people analytics team uses LinkedIn Talent Insights. This allows them to access external data like external benchmarks and trends over time. 

4. Analyze: Gather actionable insights from your analysis to prove or disprove your hypothesis

As you start running your analysis, this is your opportunity to dig into why the business problem is happening. In the case of attrition, you might look at data from your enterprise resource planning (ERP) system and exit interviews to spot correlations between one team’s unusually high turnover and the reasons employees gave for leaving. 

“Did they leave for personal reasons?” Rebecca asks. “Did they get a better job? Or maybe they didn’t like their manager.” 

These insights can help you identify possible root causes behind the business problem, which will inform your recommendations. This can be exciting, but it’s important not to race ahead before you’ve confirmed the accuracy of your findings. 

If your sample size is too small, for example, then any insights you gather may not be representative of the department or workforce as a whole. This could cause your team to waste time making recommendations that will have little impact, which could hurt your credibility in the long run. So always do your due diligence and make sure you feel confident about your findings before you share them. 

5. Communicate: Develop and share recommendations based on your findings 

Once you’ve run your analysis, it’s time to communicate your findings to your leaders. This is what bridges the gap between insights being merely interesting to becoming actionable. 

“The communication is critical to making this stuff happen,” Rebecca says. 

Think carefully about the medium you’re using to convey your findings (like an easy-to-read chart versus a complicated spreadsheet) and the language you’re using. Your goal should be to create clarity and tell a story, since not every leader will understand what the raw data actually means in terms of business outcomes. 

“There's definitely varying degrees of data literacy,” Katie stresses. 

Often, there won’t just be one singular root cause you can point to for any given business problem — most issues are more complex than that. But what you can do is identify which causes are within your company’s control, allowing you to start formulating recommendations. 

“We can say, ‘Here are a few key trends or drivers that we think are causing this uptick,’” Rebecca says. “And the next part is, ‘Based on those, let's identify what are the things we can do.’”

With attrition, people often assume that compensation is the culprit. Rebecca says it rarely is, but if there is a correlation, your team can recommend that the company reviews its compensation policy to ensure it’s still in line with the market. Or if your analysis shows that first-year attrition is particularly high, you might suggest that the company reviews its onboarding process to provide additional support for new hires. 

“We might say, ‘Okay, we need more time spent on onboarding,’” Rebecca says. “We’d recommend that's a program that we need to change or expand. Then our learning and development (L&D) team would go away and think about how to do that.”

6. Track: Monitor the results of any actions taken based on your findings and recommendations 

After handing off your recommendations to relevant stakeholders, your people analytics team still has one step left to take: Measuring the impact of any actions taken to fix the business problem. 

If the L&D team is building a new onboarding program based on their recommendations, for example, Rebecca and her team would be invested in monitoring the effectiveness of that program. 

“Say our first-year attrition was 5% this year,” she says. “And a year from now, our goal is for it to go down to 2%. Our team would be in charge of setting up those metrics and trying to track the impact of these interventions.”

Building close partnerships with teams across the business will make it easier to ensure your recommendations are being followed and that any impact they have is tracked and documented. This is what will get your team recognized as a strategic partner to the business and encourage other stakeholders to tap you for insights and recommendations. Otherwise, you might be spending a lot of time doing analysis that ultimately leads nowhere. 

“There's a wealth of data in the world, but is there any buy-in from your company to actually implement and track those changes?” Katie says. “A lot of times when I was starting out in my career, I would just do an analysis and then I'd have no way to replicate it in six months when somebody wanted to know, ‘Has it changed?’ because we didn't actually set up any controls to know that somebody adopted a recommendation.” 

Final thoughts 

People analytics can seem big and complicated when you’re first starting out. By following this six-step framework and breaking it down into less intimidating chunks, your team can get into a productive rhythm and begin identifying useful insights that drive positive change. 

As you get more confident and skilled at people analytics, you can even start to do more advanced analyses — building and testing models to predict issues like attrition before they happen. But the first step is mastering the basics. Once you’ve done that, the sky’s the limit. 

For more insights on how your team can get started with people analytics, download the full Global Talent Trends 2020 report today.

To receive blog posts like this one straight in your inbox, subscribe to the blog newsletter.

Have blog stories delivered to your inbox