Continuous Improvement to Café Services

This is the Story

Café services at my place of work are not the typical café services you may have experienced.  This is not your typical stainless steel hot food setup with bland, greasy options and the lingering smell of a fry station.  These are vibrant, open, music-filled spaces offering a high-volume, high-end, customer-experience-oriented restaurant operation.  This is where you go to relax, decompress, be with friends, and give your taste buds a foodgasm (is that a thing?).


Imagine salad bars structured to allow more than a million combinations of plates, based on the selection of proteins, dairy products, vegetables, seeds, and dressings; food stations for vegan, and vegetarian options, a grill station, international food stations, pre-packaged takeout options (for those in a hurry), invited chefs who teach employees the ins and outs of special dishes, and a desserts station packed with at least a dozen choices of exquisite “I’m in heaven” bites.  Add to that free fruit and drinks, a variety of at least twelve different brews of coffee from local and international vendors, teas, spa waters, and your typical soda fountain… all of this feeding thousands of people per day on a 24/7 operation.  (I will post later about their catering services, where I measured catering deliveries between buildings adding up to 90 and 100 miles of walking per week – I know, unbelievable.)


When I was first asked to do something in this space, I was just asked to “go help them”.  That’s it.  There was no problem statement, there was no clear business case, and no goal; it was just “Rick, they’re struggling, go see what you can do with them.”  I had never thought of café services; must admit I always took the café for granted, but I was lucky enough to connect with a few of their managers who besides being wonderful people, they were open to let a stranger into their space and look at what could be done better.


To not walk in blind, I started by looking at their data to understand peak times and traffic volumes.  We did an analysis of variance in the consumption of café products by the time of day, and another on the food items prepared by station per day of the week; looking at this data would indicate hot spots and opportunities at two levels of time resolution (day and time).  We also ran control charts to understand the variation in the mean and moving ranges of their throughput.


The Situation

The data dictated how we would prioritize the areas and we headed to where the action takes place for direct observation, map the flows of people with spaghetti maps, and start time studies at the constraint.  This is what we learned:

The Current State?  Not pretty...

The Strategy

Stabilize the system by applying short-term containments where needed.  This meant rectifying broken connections between the stations, ensuring no “inconvenience factors” were interrupting the customer experience, and engaging the team by teaching them the fundamentals of workplace organization (you may call that “5S”).


Once we stabilized the system, we sought to minimize the throughput times and structure work around the constraints.  We brought with us a Safety Engineer and an Ergonomist to identify potentially unsafe conditions, certify we exceeded the expectations of the Health Code, and flag any potential ergonomic risks.

We standardized all inventory replenishing of consumables

We started by defining cut-off times to restore inventories of consumables: tea packets, fruit baskets, cups, and lids, refill the honey containers, and sweeteners, and restock ice levels in all soda machines.  The data suggested the timeframes.


We standardized the method to place orders

We looked at the congestion of customers in certain areas and applied the fundamentals of “Drum-Buffer-Rope”.  We learned there were two methods of placing orders, some would "tell the cook", and others will fill up a paper slip.  This little variation was pushing people away.

People waiting for their orders in the immediate area created artificial congestion.  Many were on their phones leaning against a wall, others leaning against the Salad Bar, blocking access to trays... it was messy.

We asked a few customers who stopped and turned around to leave.  One of them said: 

“This feels too congested… stressful… I mean who wants that?  I need my lunch break to relax.”  

And they would go somewhere else.

We organized the cooking workspaces

The grill station had no designated space to set meals in progress, the cooks were overreaching to hand out orders, and they were trying to make use of visual management, but it was not consistent.  The layout was not optimal.

We mapped the process flow required to run the Grill Station, but with the goal of maximizing the customer experience

Data Analysis showed other constraints at the Grill Station – We fixed those too.


(Yellow highlights show the impact of an inrush of orders, affecting the time to cook and forcing the addition of an extra person to help put orders together.)


In a matter of a week, we turned things around.  

Who would've thought?  Poor methods at the Grill station, next to the main entrance of the cafe, were directly affecting the customer experience of the whole cafe operation.

We fixed that and everything else fell into place.  Not only the TPT improved but notice the higher density of orders in the new current state.

A customer said:

"I don't know what has changed, but it does not feel crowded, I like that I can walk around.  I even got my burger in less than five minutes!"

Take it to the Bank baby!


Our Service Provider told us that by improving the customer flow, increasing the volume of transactions, reducing the waiting time, minimizing the rework and rejects because of “wrong orders”, and bettering the customer experience, they anticipated an increase of $300k in revenue by the end of that year, from only THAT cafe.   Look at the improvements:

Results by the Numbers

The volume of transactions at the cash registers increased by 102% (applying 5S only) and by 142% (once we maximized the constraint at the Grill Station).

Lunch Order TPT at the Grill reduced by 30% (anytime you get your lunch faster is a good thing).

Reduced the wastes of Motion and Transport in such a way that the cook could serve 9 of every 10 people accessing ingredients from a 4-foot radius at the workstation.

Salad Bar setup time was reduced by 7%, giving back the crew 20 minutes; those 20 minutes per setup add up to a lot of minutes a year.