Nestlé nestles down to HR analytics

Profile picture for user jmilne By Janine Milne October 21, 2015
Analytics is helping food giant Nestlé uncover the secrets behind high staffing attrition rates.

Analytical thinking
Analytics isn’t just a silver bullet…There was a perception that l would answer this problem and it will be straightforward and it would be just one thing.

The reality, according to Katy Bowers, UK senior HR data analyst at food and drink giant Nestlé, is that using analytics is rather more complicated. She told delegates at the CIPD London analytics conference:

People are complex. There isn’t one magic answer; it’s a combination of factors and it’s how they play out and interact with each other.

The problem Nestlé was looking at was rising attrition rates among the UK’s 9,000 employees and how to link that with its engagement strategy.

Intensifying competition and the influx of millennials had changed the pattern of attrition in the company, making it very different from when Michael Cox, head of HR analytics, joined the company a decade ago:

If anything, attrition was the opposite problem of no one ever left …clearly when we look at it now, we have high attrition.

Cox heads up a four-strong HR analytics team looking at this and other HR issues, but alongside the central team, Bowers notes that it’s key to have a corporate sponsor, someone with clout in the organization, who can help keep them on the right track and fight their corner. At Nestlé UK, that person is the head of HR operations.

Running analytics takes a lot of effort, so it’s important to have it linked tightly to strategy to ensure this investment isn’t wasted, notes Cox:

Particularly as we move into more areas of predictive analytic going forward, but even with diagnostic analytics, it takes time, it takes investment...there has to be a really compelling business case and there needs to be a tie back to return on investment – in terms of skills in terms of analysis, what’s it going to deliver back.

Data everywhere

But alongside the team and a firm grip on business requirements, there has to be data to work on – and there’s plenty of that in a company the size of Nestlé. Bowers explains:

We had about 500,000 pieces of information in total, touching all our different things technologies. We had information from surveys, we had core SAP data, we had payroll information, recruitment data, talent information, succession planning information. Literally every form of spreadsheet or tool, we touched it in one way or other.

It was a challenge to “join all the dots” between the different data sets. But, while it’s important to have good data – and the team spent time upfront ensuring the data was in a good state – Bowers is keen to point out that it doesn’t have to be in pristine condition:

I guess our message on this, certainly in our experience, is that it doesn’t have to be 100% right…Do you know what – 80% to 90% is good enough for what we’re trying to do. …Whether the attrition rate is 10% or 11%, your actions are the same.

The team used Minitab to analyse the data and make correlations between turnover rates and employee attributes to build a profile of the people leaving the company.

Drilling down in the data, the team found that the head-office workers had significantly higher attrition rates. They determined that the main problem was with the Nespresso division, which had a different business model from the main business.

One of the things they found was that women were leaving than a higher rate than men, regardless of their level or performance achievements.

They surmised that high levels of attrition rates of women workers was down to them not returning after maternity leave. Yet, in reality, notes Bowers:

I was able to show was that even if we excluded data about people on maternity leave, it didn’t change things; the problem was still there.

Then the team looked at whether succession planning was at fault and whether there was a problem with a lack of transparency about the next steps up the career leader. Another hypothesis was that perhaps that there were problems with performance ratings. But again, the data said, no.

Such negative results are actually a good thing, notes Bowers:

There’s a value in not finding something. Sometimes that’s as useful to the business in terms of shaping the direction.

It’s all about narrowing down the variables and “having that scientific mindset” to being able to disprove what on the face of it seems a logical deduction, agrees Cox.


The team were able to make correlations between turnover rates and employee attributes and found that there were two distinct leaver profiles. It also uncovered five reasons why people were leaving: remuneration, leadership, recruitment and induction, leadership and culture. Armed with this information, Nestlé has been able to changes in all those areas to try and improve engagement and reduce leaver numbers.

For example, in the area of recruitment, the data showed that they needed to start their induction process earlier. Rather than their first day on the job, this process should start at the pre-boarding stage, when someone walks into an assessment centre of apply online. Cox expands:

That’s triggered a piece of work around have an interactive portal in place to engage with people and allows them to chat with colleagues and ask questions before they start.

It is also putting its exit data to work, rather than just leave it languishing in a database and doing nothing with it.
Bowers emphasizes that it’s important to use a number of techniques and to look at the bigger picture, rather than rely solely on what your analytics tool tells you:

As you can imagine, we’ve got so much data in HR and it’s only getting bigger and bigger, the demands are getting greater and this is why we take a step back and focus on what matters.

The way we get to that isn’t just through analysing the data, but also through that internal partnership with our sponsor and HR community out there to get there knowledge and their experience. And we also look at what’s going on in our external market….If you have just one of those three things, it doesn’t work.

It’s all about context; data alone will not provide the answers needed.

The analytics team also need to be able communicate their findings. All this hard work would be worthless unless people believe what they are told, says Cox:

We’d done our homework and we really understood the data and we’d really analysed things and proved things statistically and it gave us much greater credibility, not just within HR but at business level.

Having that credibility meant that the team felt confident in their abilities to influence and negotiate with the organization. Their work doesn’t stop there. It’s important not only to make suggestions, but to follow up and see they’ve been acted upon. Cox swears by the mantra that:

Any powerful idea is absolutely fascinating and absolutely useless until we choose to use it.