Ethnographic Report on Potential Users’ Restaurant Behavior

Luke Zaccaro
MI 841

Michigan State University

Fall Semester, 2016

Executive summary

Observational research was conducted to inform the design of, and determine the best target audience for, an app designed to streamline the ordering and payment processes of eating establishments.  The goal of this app is to meet the challenges of customers who want a sit-down meal with high quality food, but want it to be a faster and more efficient experience.

Direct observations were made of customers’ restaurant behaviors over a 95 minute period.  Special attention was paid to the amount of time it took them to order their food, and their activities prior to making menu decisions.

Key Findings

  1. Group size determined how distracted customers are, which increased the amount of time it took them to decide on an order.
  2. The greatest potential for (and presumably motivations for) saving time in the ordering process was among parties with small children.  They took the longest to order their food, and had the highest level of distraction and interruptions.
  3. An app that allows customers to pass data directly to cooks bypasses a server and can reduce mistakes


I observed customers enter, order, and eat/drink over a 95 minute period in a small cafe near my home.  I completed my observations during the lunch period from 11:00am to 12:35pm.  I recorded a total of 8 parties totaling 18 people in that time.  The cafe is an upscale suburban restaurant, and the food is high quality.  The menu is standard cafe fare - eggs, hashbrowns, etc for breakfasts, burgers, salads, etc for lunch and dinner.  I moved to this town about a month ago, and I have already eaten there a few times, so I am familiar with it. It can get fairly busy, and demand for tables is moderate to high on weekdays.  The dining room is L-shaped, and I was fortunate enough to be seated in a corner where I could easily observe half of it.  I recorded the following information about every party that entered my half of the restaurant:Cafe Layout.jpeg

  1. Time they arrived at their table
  2. Party size
  3. Ages and genders of adults
  4. Presence of children (no identifying information)
  5. Seating Location
  6. Activities before order was taken
  7. When the order was taken
  8. Activities while waiting for the food to arrive
  9. When the food arrived
  10. Activities after the food arrived
  11. When the party left the restaurant (3 groups stayed longer than I did, and so I have no data on when they left)

Party Name

Group Size

Age and Gender


Distraction level before ordering 1-10

Time to order

Activity while waiting

Grumpy Gus


M, 65


0: Looked at menu


reading something

Suburban Couple


M/F, 45/45


3: Looked at menu, talked a little


on phone

dying marriage


M/F, 45/45


1: Looked at menu, did not talk


occasional conversation

we just retired


M/F, 50/50


2: Looked at menu, then talked


Lots of conversation, but the guy keeps looking around the room

Political Brunch


F/F, 60/60


5: Alternately discussed menu and current politics


constant political conversation.

White Widows


F/F/F, 60/60/65


6: Deliberated, conversed, asked for other menu, rambled


quiet, polite conversation over coffee

Girls interrupted


F/F, 30/30


8: Forgot to look at menu, looked, forgot, etc. Distracting Toddler


Adults tried to converse while toddler periodically interrupts for attention.

3 generations


F/F, 60/30


6: Looked at menus and conversed - kids played with some toys but interrupted occasionally


Kids grew restless - bouncing, standing, etc. Adults worked hard to occupy them

This summary table shows some of the salient data from the parties I observed.  I gave pseudonyms to each party, and documented their activities before and after ordering food.  I’ve shown the parties with kids in red, and the parties without in blue.  Bold type font indicates that party members arrive separately, and normal indicates that they arrived together.  The table goes from a group size of one, with “Grumpy Gus” at the top, to a group size of four, with “3 generations” at the bottom.  There is a direct correlation between the size of the party, and the time that party takes to decide their order.  I will discuss the “distraction level” later.


Time to prepare the food

Even with a small sample size of just 8, there were a number of clear trends that I discovered.  First, I will start with the results of how long it took the cafe to bring food to the participants’ tables once the order was placed.

From Order to Food.jpg

The cafe got food to the tables in a mean time of about 14 minutes and 34 seconds.  If I remove the outlier time of almost 29 minutes, (I’ll attempt to explain that one shortly) the mean time for food delivery is just over 12 minutes.  Consistency is quite good - with the exception of the one clear outlier, deviations from this time were all less than 3:30.  While I could not see the menu items that most people ordered, “Political Brunch” happened to be right next to me, and I saw that they ordered a standard breakfast of eggs and hashbrowns.  I know from my experience as a server that these items can be prepared in under 8 minutes.  I also know from my experience, however, that it is common to forget to put an order through to the cooks immediately after taking it.  I have no proof, but I strongly suspect the server saw that this particular table didn’t have their food after about 15 or 20 minutes, knew that something was wrong, investigated, became aware that the cooks never got the order, and then immediately alerted the kitchen, resulting in food delivery of almost 30 minutes.  I saw this kind of thing happen quite frequently during the brunch/breakfast rush during my shifts as a server.

Note: I omitted “We Just Retired” from this list, because I did not notice when they got their food.

Group size and ordering speed.

While food preparation time on the part of the restaurant was consistent, the amount of time it took customers to decide on what to eat was not.  The single largest factor for determining how long it took to peruse the menu was the size of the group.

Group size and Order speed.jpg

Here, the table is arranged with the y axis showing the amount of time a table took to order, and the x axis showing the size of the group, increasing from left to right.  The larger the group, the longer that group took to order.  


I also took my original notes on the activities and demeanors of the participants, and created a “distraction rating” for each table.  This is a completely subjective rating of 0-10, with 0 being not distracted at all, and 10 being very distracted.  I define distraction simply as any activity other than reading the menu.  I attempted to assign this rating to each table for the time from entry to ordering.  Each table’s distraction level can be found in the summary table, but I’ve graphed the relationship between distraction level and the time it takes a table to order their food.

Group Size and Distraction.jpg

Participants are again arranged from left to right in increasing order of group size.  A “T” indicates that the adult participants arrived together, and an “S” indicates that they arrived separately. (“Grumpy Gus” was by himself.)

While there is variation between groups of the same size, it is clear that larger groups are more prone to distraction than smaller groups. Arriving together or separately also tends to increase distraction.  Also, the presence of children, not surprisingly, can increase distraction level as well.  “Girls Interrupted” and “3 Generations” were the two most distracted parties, and were also the only parties with children.  This distraction also affected how long it took them to order their food, as shown below.

Presence of Children

Entry to Order.jpg

Here, parties with children are again shown in red, and those without are in blue.  The presence of children clearly increases the amount of time it takes for that table to place an order.


Restaurants are relatively consistent with how long it takes to get food out to the tables.  Cooks and servers generally want to be as efficient as possible in their jobs.  There is a great deal of variation in customer behavior, however - particularly in how long it takes them to decide what they want to order.  The key determining factor in how long it takes a table to order is how distracted they are.  This distraction is increased by group size, the presence of children, and arriving separately.  These are the areas any tool that seeks to streamline the ordering process should focus on.


  1. Why does arriving separately increase distraction level?  I suspect this is because parties that arrive together (usually couples) have already been talking and interacting prior to arriving at the restaurant.  They are not tempted by social norms to engage in greetings, pleasantries, or “catching up.”  Parties that arrive separately, however, may find it a bit rude to say hello, and then promptly ignore each other to read the menu before picking up the conversation again.  These kinds of users would not benefit much from an app that allows you to order ahead of time, as their pace is generally more leisurely.  However, if they paid for their meal (and tipped) in the app, then they would be able to leave without going through the long and difficult manual process of paying separately, which often requires many trips by the server, and long cashout procedures.

  1. If the motivation for eating a meal at a restaurant is social and/or recreational, then longer ordering times are probably not a problem.  If, however, the motivation is to get a meal into your children without having to cook and clean, then you want the process to be as fast as possible.  It is much easier to occupy young children at home than in a public place.  (It’s also easier to administer consequences for negative behavior at home)  So why do these families not simply go to a fast-food restaurant?  They may want higher quality food that is typically not available in faster establishments.  These are the perfect users for an app that allows customers to order food before arriving.

  1. Mistakes are made by servers and/or cooks.  One patron, “Grumpy Gus,” sent his food back because there was some ingredient in it that he asked be left out.  I heard the server say "I'm sorry - I told him to leave them off.  I'll put in another one for you.”  I don’t know who made the mistake - the server or the cook, but as a former server, I’ve definitely blamed my mistakes on cooks before.  The customer doesn’t tip the cooks, and generally never sees them.

    The presumed mistake with “Political Brunch” had no consequences - the table of older women had arrived separately, were apparently retired, (late morning on a weekday) and gave no indication of being in a hurry - they had been there almost 80 minutes when I left.  If, however, the server had made a mistake with one of the parties with children, it would have caused much more frustration on the part of parents to add an additional 20 minutes to their burden of occupying their kids.

    While mistakes are inevitable, an app that allows people to order their food before arriving would at least eliminate ordering mistakes by the server - the order would get sent immediately to the cooks, cutting out one participant in this food-based game of “telephone” that we all play in restaurants.

  1. While distraction level and ordering time for the single man were low, he still had to wait almost 15 minutes for his food.  Someone like this could also benefit from ordering ahead of time.  I believe he was there on a lunch break from work, as he was dressed in professional clothing.  If he ordered from his desk, he could have a meal (higher quality than fast food) waiting for him when he arrived.  

Design Recommendations


This app must be able to do the following:

  1. Users peruse a restaurant’s menu ahead of time and select their desired item, along with all possible options. (fries vs salad, etc.)
  2. Users will have to pay for the food through the app (tied to paypal, credit card, etc - like Uber) before arriving if they want the restaurant to start making the food before they arrive.
  3. Cooks will be automatically notified when an order comes in, and begin preparing the food without being given an order from the server.
  4. Users must be able to create or join parties to group their orders
  5. There must be a “table management” system to reserve tables for users who are about to arrive.
  6. Users can request things like extra napkins, refills, or desserts right from the app.  This allows the server to immediately address table needs without having to be flagged down and asked.
  7. Participation is completely voluntary - restaurants should be able to embed this app into their current systems without requiring the customer to do anything differently.

Target Audience

  1. Families with young children.  
    This is the group that would benefit the most from this tool.  They would save time by ordering and paying from their device, and time is their scarcest resource.  Arriving at a restaurant to a meal that is ready or almost ready would be extremely valuable to them.  
  2. Larger parties of separately arriving people.  
    This group would also benefit from this tool. Using the app, one member, “Organized Oliver,” would create a party, give it the name “Oliver’s Twisted Table,” and then other users could join it after using the app to select the restaurant.  Orders could be placed and paid for before arriving.  In my experience, members of these parties may order multiple drinks throughout an evening.  With this tool, they could do that right from their phones, so they can spend more time conversing with each other, and not the server.

My raw observational data sheet can be found here: