Sign up for content marketing news and tips delivered every Tuesday.
How Consumers Search For A Perfect Meal
How do consumers find the perfect restaurant? Word of mouth plays a role, but there are other elements at play. How important is your brand? Is value more important than quality? What role does content play in this quest? Here I take a look at the term “restaurant” to show what consumers value when searching for a good place to eat.
In this data set there are eleven distinct high-level categories of behavior. The categories are sorted by volume (the first number reflects the number of keyword phrases in that group).
- 22 Informational – 65,650,450 searches
- 308 Location based – 13,577,700 searches
- 76 Type of restaurant – 4,261,540 searches
- 51 Quality of restaurant – 1,835,520 searches
- 36 Restaurant business – 1,299,270 searches
- 150 Restaurant by brand – 1,236,530 searches
- 81 Request for content – 850,020 searches
- 42 Value (cost reduction) – 500,660 searches
- 12 Industry events – 376,600 searches
- 4 Restaurant directory – 218,000 searches
- 1 Software tools – 2,400 searches
When you examine these categories you are struck by the specificity of intent here. Generally, there are large numbers of vague informational keyword phrases being used. In this case, there are just 22 of them, but they account for 65M searches a month. The second important item to note are the top three themes that dominate restaurant search behavior: location, quality and type. From an information architecture perspective these three concepts, which represent almost 30 percent of consumer traffic, should be infused into website copy. The probability that a consumer will construct a [quality] [location] [type] query is very high.
Several of these high-level categories have sub-categories that provide useful information about consumer intent. This expands the unique categories to twenty seven. There are, for example, three categories of behavior for content, and five categories for the Type of restaurant.
- Information – 65,602,880
- Location – 13,538,000
- Type by nationality – 2,380,480
- Quality using Best – 1,251,600
- Brand – 1,211,340
- Business products – 1,043,370
- Type by food – 901,190
- Type by style – 733,390
- Industry events – 376,600
- Content (reviews) – 352,120
- Quality using Top – 345,720
- Content (guides) – 277,180
- Value (coupons) – 266,790
- Business – 255,900
- Quality – 238,200
- Website directory – 218,000
- Content (menus) – 168,020
- Type – 136,480
- Value (vouchers) – 126,080
- Type for delivery – 110,000
- Value – 107,790
- Content – 52,700
- Information (names) – 44,200
- Location by country – 39,700
- Brand by location – 25,190
- Information by zip codes – 3,370
- Software tools – 2,400
When you look at consumer intent by category, you can make assumptions when developing your website architecture. For example, many more consumers are interested in reviews rather than looking at menus by a two to one margin. Both content types are important, but from a volume perspective restaurant reviews should get top-billing.
Let’s take a look at the categories in a little more detail.
Brand: There are 1,211,340 searches a month for restaurants by brand name alone (no location was specified). Interestingly, the search for a brand by location comes in on the low side at 25,190 searches a month.
Business: As a restaurant business owner these searches are of no value to you. The majority of the traffic is for products and services (1,043,370 searches), while the rest (255,900 searches) are about franchises and for sale opportunities. The majority of the traffic for business products is searched for with just a handful of secondary terms. This makes it straightforward to manage where your ads should not be displayed. These terms in order of importance are:
Content: These are, for the most part, three very specific content requests. Reviews are the top-dog in terms of volume, but this volume is clustered in just 10 keyword phrases. Guides show up in 53 different queries, most of them city-based. Most of the request for menus is by national cuisine (Chinese, Italian and Mexican). There is very little traffic for menus from brand name restaurants.
- Content – reviews (352,120 searches)
- Content – guides (277,180 searches)
- Content – menus (168,020 searches)
- Content – non specific (52,700 searches)
Events: There are 376,600 searches a month for restaurant week. Coming in at four million searches a year, this looks like a microsite opportunity worth exploring when this event occurs in your city.
Information: Not surprising, the majority of this traffic is on a single term: restaurant(s). It’s hard to divine user intent in this traffic. It could be businesses looking for services, or a mom looking for a birthday party restaurant. There are a number of queries looking for a list of restaurant names by nationality. Are these queries coming from someone looking to start a new restaurant, or is this an unusual way to search for a place to eat? I suppose you could target this traffic to see how it converts.
- Information – vague intent (65,602,880 searches)
- Information – request for list of names (44,200 searches)
- Information – request restaurants by zip code (3,370 searches)
Location: With 13,538,000 monthly searches, these location-based searches are the second largest behavior category. This group is looking for a restaurant in a city without mentioning any other attributes, such as food type, content or quality.
Quality: Consumers are certainly interested in a quality eating experience, and they use relatively few adjectives to describe the restaurant they are looking for.
- Quality using the term best – (1,251,600 searches)
- Quality using the term top – (345,720 searches)
- Quality, various terms e.g., good, famous, 5 star – (238,200 searches)
Tools: There are just 2,400 searches a month in this data-set for calorie counters. At this level you can’t really be bothered with such a tool. Perhaps if you are offering fast foods this becomes a more useful option.
Type: There is a lot going on in this group, and it provides a restaurant owner the option to construct landing pages with ad copy that reflects the probability that consumers are searching for your restaurant in five different ways. The probability that consumers will search by national cuisine is very high – the probability that they will search by delivery is almost twenty times lower.
- Type by nationality e.g. Chinese, Mexican – 2,380,480 searches
- Type by food e.g. seafood, burger, pizza – 901,190 searches
- Type by style e.g., family, romantic, buffet – 733,390 searches
- Type, various one-off terms e.g., gluten free, waterfront – 136,480 searches
- Type, for delivery – 110,000 searches
Value: Cost is important in this category. The majority of consumers focus on just two terms: coupons and vouchers. When consumers search for coupons, you do see some request for coupons by brand name. When they search for vouchers, none of the consumers specify a brand.
- Value – coupons 266,790 searches
- Value – vouchers 126,080 searches
- Value, various terms – e.g., cheap, kids eat free 107,790 searches
Website directory: 218,000 consumers a month search for directories of restaurants. There is certainly enough traffic here to justify registering your restaurant with every local and hyperlocal directory you can find.
Restaurant Search Behavior Model
It’s always useful to look at term density in the dataset, as it provides more insight into what customers value. In this case the top term is best. This would suggest a great landing page opportunity e.g., “the best seafood restaurant in Boston.” In this label you have captured the three main themes identified earlier: location, quality and type.
Stop Words: If you have been around search for a while, you know that stop words (a, in, the, of, for, be) are supposed to be of little value in search relevancy. However, if you look at the list of secondary terms below, you see that the term “in” has the fifth highest density. In this case, it is very valuable because it is a geo-indicator. You see a lot of traffic using the following two forms:
- Best in [your city]
- Restaurants in [your city]
Since the best search result is an exact match with a consumers query, the term “in” becomes important because:
- It’s used often
- It’s a geo-indicator
- It’s important in an exact (word-for-word) query match
By the way, the term “near” plays this roll as well, as in restaurants near Kendal Square.
When you examine the rest of the top 25 secondary terms you see several themes reflected in the list. These are listed roughly in order of value. The first six provide insight into what is valued by consumers, and should govern how you develop your website architecture and page copy.
- Looking for a quality dinning experience
- Looking for restaurants by national cuisine
- Looking for content (reviews and guides)
- Looking for a restaurant by food type (seafood and vegetarian)
- Looking for a restaurant by style (family and romantic)
- Looking for value (coupons and vouchers)
- Looking for directories (city cheat and table-table)
The secondary terms in the following table (in the red font) are off-topic, and are of no use to you. In fact, if you talk about the state-of-the-art restaurant equipment you have installed in your restaurant, you may well attract traffic looking to buy stoves and chairs.
So, what does this analysis do for you as a restaurant owner? Let’s list some of the more important items to think about.
- At the highest level, consumers are interested in location, quality and type. These themes should underpin your website copy.
- Consumers are much more interested in quality than value by a three to one margin. The terminology they use (best and top) and the most requested content type i.e., Reviews supports this observation.
- When consumers search by brand, they already know about you. If they have never visited your restaurant before you should support this group with reviews and location information.
- Many consumers are cost-conscious, and search for discount coupons. What was a bit of a surprise in the data was how often the term voucher was used. I think this reflects the relationship that hotels have with the local restaurant community. I don’t think I’ve ever received a coupon from the concierge, vouchers yes, coupons no.
- Last but not least, you have the option to provide potential customers with three distinct ways to find your restaurant. Statistically large numbers of consumers search by nationality, by food type and by the style of the restaurant. This behavior suggests a multiple landing page strategy.
The last point about the style of restaurant is important. What if you are “family restaurant?” Does search behavior differ when searching for a restaurant when children are factored in? It does and it doesn’t. Many of the categories of behavior are the same, but with dramatically different volume. For example, there is also almost no interest in reading reviews in this group, and brand searches tops the list. Next month I will develop a family restaurant search behavior model, and contrast it to this model to show that behavior can be different depending upon the type of restaurant consumers are looking for.
This analysis is high-level, and it is very useful to understand what is going on at the industry level—the dominant themes and so on. However, most restaurants are not one-dimensional. They have multiple characteristics such as in this tag line: “Brazilian steakhouse specializing in romantic dining.” My advice here is to go the extra mile to understand the behavior associated with consumers searching for a romantic restaurant. No doubt, the behavior will differ from this analysis.
As in politics, all restaurants are local. If you have not claimed your business in Google’s, Yahoo’s and Bing’s local search services, you should do so. You will not automatically get added to these local search indexes. There are local search requirements for being included. Make sure you understand what they are.
The data used in this analysis was extracted from AdWords.
Some opinions expressed in this article may be those of a guest author and not necessarily Search Engine Land. Staff authors are listed here.