You are here:

Seven things to consider if you want to run a controlled experiment in your staff restaurant

Healthy eating in the workplace

About this article

  • Published

    05 Jan 2017
  • For regular updates

Are you thinking about introducing an employee wellbeing initiative at your company? If so, are you planning to measure the impact to ensure it delivers the benefits you hope for?

Many companies have made changes in the workplace to support employee health but in our experience, few of these have been rigorously evaluated to assess their impact. This particularly applies to healthy eating initiatives at work. There is very little evidence available of what really works to improve employee diets.

If your business was launching a new product, you would be sure to measure the results. Employee wellbeing is also a vital investment, so it’s equally important to ensure you’re getting a good return for your efforts.

We’ve embarked on a huge experiment across many company sites to test various ways to encourage healthy eating and we’ve learned a lot already. So to help you measure the impact of your employee initiatives, here are seven key factors to consider:

The points are illustrated throughout with learnings from our multi-company experiment on healthy eating in the workplace.

1. Nominate a head researcher

Someone needs to oversee the experiment, preferably with experience in conducting research.

Depending on your objectives and budget, this might be drawn from your own organisation or an independent partner, such as an academic institution or a research company. An impartial person, with no vested interest in the outcome, will help to minimise any potential for bias. An expert partner would also help ensure your experiment is credible and robust if you plan to share the results.

We partnered with the Behaviour Health and Research Unit (BHRU) at the University of Cambridge.

2. Decide what to test – one thing at a time

There are many areas you could work on e.g. diet, physical activity or mental wellbeing. However, if you introduce lots of changes at once, you won’t be able to isolate which is having the most impact. It’s better to move one step at a time, checking and learning as you go.

To help prioritise, it’s worth considering any existing evidence e.g. from published papers, case studies or the experience of your own employees in their previous companies. Always keep your end goal in mind, i.e. the ideal outcome for your employees.

To help us decide what to test, our partners at the BHRU had reviewed all available evidence to identify which interventions were most likely to bring about positive dietary change in the workplace.

Most of the studies had only been conducted in small or ‘laboratory type’ settings (not in a real cafeteria) highlighting the need for new research but at least providing a steer on what we should test.

Having reviewed the evidence, the three interventions we decided to test first were:

  1. Labelling – clear, simple energy (kcal) labelling at point of food and drink choice
  2. Availability – increasing the availability of lower vs. higher energy (kcal) food and drink
  3. Sizing – reducing the portion size of certain food, drink and/or tableware items

3. Get all the players on board

A successful wellbeing programme can’t just be mandated from the top; it needs buy-in from lots of people to be executed well.

Equally, an experiment can easily be sabotaged by someone who wants it to fail, for instance by spreading rumours, so a stakeholder management plan is essential.

You will need to consider who are the key players, what might their initial reaction be, what extra work might it entail for them, what impact might it have on their own objectives and what benefits could it deliver for them over time?

This might include your suppliers too, for instance many companies use contract catering companies.

It’s important to ensure you have the right support and networks in place before you commit to anything. You will probably need to adapt your messages and language to different audiences. What wins over an HR Director might not work for a Head Chef.

You also need to listen effectively, take on board people’s valid objections and be prepared to adapt your approach without sacrificing your end goal.

Our early success has been the result of extensive groundwork, explaining our intentions, gaining support from CEO level to the restaurant staff and many points in between. It’s been a working collaboration between the host companies, their catering providers, IGD and our research partner, the University of Cambridge.

4. Getting systems in place to measure change

Any experiment requires data, so you need to consider how to measure changes in employee behaviour. This will depend on what your company considers a success e.g. reduced blood pressure, weight loss, reduced sickness.

To measure the impact of any change, you’ll first need to consider your starting point. In scientific terms, this is called ‘baseline’. You may be able to measure ‘baseline’ from data you already collect e.g. cafeteria sales, employee health records etc. or you may need to introduce a new source of data.

For our experiment, we have been measuring total energy (kcals) per person consumed. Participating sites are sharing cafeteria sales along with calorie or recipe information for all the food and drink items they sell. Our research partner is comparing this data, before and after each change.

5. Don’t let people know they’re part of an experiment

Would you act differently if you knew you were being watched? Many people do, whether consciously or unknowingly and this is known as the Hawthorne Effect.

Depending on the changes, it may be necessary to communicate some information to your employees to avoid misunderstandings but take care what you say and when.

It’s all in the balance, providing enough information to keep people happy and engaged but not too much to alter their usual behaviour, especially during the baseline period.

6. Minimise other changes

Once your experiment is underway, keep other variables to a minimum. As with testing too many things at once, if you introduce another change part way through, it is likely to disrupt your results.

When you’ve decided how long your experiment will run for, consider what impact other factors could have on your results e.g. time of year, special events and holidays. For instance, more of us reach for that extra treat in the run-up to Christmas whereas the gym is always a little busier in January!

Choose a period where special factors are minimised and then however small it might be, consider postponing that new campaign or initiative until you’ve completed your experiment.

Of course, sometimes things happen beyond your control, so take note and factor these into account when you come to evaluating your results.

We asked the workplaces participating in our research to avoid things like ad hoc promotions in the cafeteria, changing the layout or introducing new food and drink items while the experiment was underway.
With the help of the host companies and catering teams, we also monitored compliance to check that everything was running as planned. There were some issues such as equipment malfunctions and free food giveaways but our research partner is accounting for these when examining the data.

7. Follow up with a staff survey

So, you’ve finished collecting data, now it’s time to see what your colleagues think.

How did they feel about the changes that took place? Would they like them to remain permanent? Has it triggered any ideas for more changes?

It may seem like you’re opening the doors to criticism but morale is an important factor of any wellbeing programme and it’s important to build a climate of trust. Listen to suggestions and be prepared to explain the rationale behind any changes. This all serves to make your wellbeing initiative even better next time round.

We’ve asked for feedback from around 6,000 employees that were involved in our experiment with a very short survey. To make it accessible for everyone, people can complete this online or in their staff cafeteria.

The benefits of rigour

If all this sounds like hard work, then remember the return you get from measuring and evaluating activities elsewhere in your business. If you believe the wellbeing of your people is vitally important, it’s worth addressing this in a disciplined way to ensure you put your efforts into the right places and improve over time.

Would your company like to take part in ground-breaking research to help your employees eat more healthily at work?

If you would like your company to help lead the way in healthier eating at work, we’re inviting workplaces with 350+ employees and a cafeteria in the UK to take part in 2017. Email us at and we’ll be in touch with further details.

Other ways to get involved

This ground-breaking research in the workplace forms just one strand of our wider Healthy Eating programme. We’re also working with industry to help shoppers make the best use of nutrition information on labels as well as helping to inspire more businesses to make their products healthier.

Find out more and register for updates by visiting

More information

If you would like to receive regular information about the Healthy Eating programme click the button below to subscribe to this newsletter.

IGD's Healthy Eating programme has been co-created by industry and IGD and consists of three parts:

  1. Nutrition information on pack
  2. Healthy eating in the workplace
  3. Reformulation

If you'd like to receive occasional email updates about our Healthy Eating programme, please register your details here.