Giving Customers What They Want: A Search Behavior Analysis
Improving website performance is hard work, and SEO alone will not get the job done. The following analysis of human search behavior is the first step in developing a world-class website strategy. Top performing websites have most of these traits in common:
- They understand in detail human search behavior.
- They have strategically invested in information architecture.
- They have a commitment to develop and deploy high-quality content on a scheduled basis.
- They understand the role quality visual design (UI) plays in successful user experiences.
- They believe in human factors, and conduct usability tests.
- They don’t let technology impact products and services in a negative way (gratuitous use of web 2.0 gimmicks).
- They have high engineering standards, and validate their code before shipping.
- They understand that SEO page markup has to be based upon quality content, not gimmicks.
- They understand technically how crawlers and search technologies impact content find-ability.
- They understand that a first-page search engine ranking has more to do with high-quality content, and a superior user experience.
You should keep all these factors in mind when developing your website strategy. However, here I will only focus upon item number one, which is how do you figure out what your customers are doing when they are looking for your services.
Search behavior analysis
Choosing keyword phrases randomly, based upon volume or Google’s suggestion can lead to success, but you never really know until you try the phrases out. This costs time and money. A better way to insure success is to take a step back, and look at the total search experience as it is reflected in an AdWords data set. This means manually reviewing each keyword phrase, and classifying it to one of several behavioral categories—usually less than ten.
A quick look at the keyword phrase “home improvement” shows that nearly 10 million people a month search for home improvement products, services and information. When you examine the data to see what is going on topically you can identify nine categories of distinct search behavior. They are looking for:
- Home improvement products: 1,477,100
- Home improvement by specific project: 1,039,700
- Home improvement by quality and value: 274,860
- Home improvement company by Name: 208,890
- Home improvement companies in general: 199,040
- Home improvement TV / Media: 60,420
- Home improvement advice, blogs & reviews: 88,000
- Home improvement projects in general: 46,490
- Home improvement financing: 6,900
Let’s look at each of these categories more closely.
Home improvement products
The largest category (1.48M) is informational in nature. In these searches the user is looking for product and information using vague terminology. They use just four secondary terms to modify the primary keyword phrase. In order of importance (volume) the terms are:
Home improvement by specific project
The second largest category (1.04M) is users looking for information so that they can learn about or transact for services for a very specific project. There are ten secondary project related terms in this group, with the top four accounting for the lion-share of the traffic. In order of importance by volume:
The remaining six project related terms account for about 5% of this categories traffic.
6. Mobile home
Looking for products and services by quality and value
The third largest category (275K) is users who are concerned with quality and value. Hands down they are more concerned with quality than they are with value. The important observation about this category is that the lexicon of terms being used is small, and they should play a role in ad copy and page markup. The secondary terms in order of importance are:
Looking for a company
The fourth and fifth largest categories (209K and 199K respectively) are users looking for a company by name, or looking for a list of companies. Users searching for a specific company by name is the larger category. A review of all the company names show that these firms fall into two major categories:
- Home improvement stores selling products
- Home improvement contractors
When users does not yet have a specific company in mind they search for company related information using more ambiguous secondary terms such as:
The Home Improvement TV series
This is the fly in the ointment. The sixth largest category (165K) is users searching for information about the Home Improvement TV series. This category needs to be understood so that the dominant secondary terms can be identified for exclusion from web page ad copy. The dominant secondary terms in this category are:
The important issue here is that a company who develops and markets “how to” videos and DVD’s for the DIY community will likely get traffic that they are not really interested in.
The seventh category (88K) reflects users early in the search cycle. They are in research mode, and are looking for ideas and information. These terms represent rich content opportunities. The dominant secondary terms in order of importance are:
- How to
General home improvement projects
The eighth category accounts for 46K search a month. These searches are informational in nature, somewhat vague and cluster around three broad terms that have several meanings:
The final category (6.9K) is transactional in nature. There is very little traffic but it is valuable traffic because searchers are looking for ways to finance their home improvement projects. The top terms are:
Observations about queries
It’s useful to remember that 71% of all consumer search terms are nouns, and 7% of search terms are adjectives. This means that nearly 80% of all queries are noun-noun or adjective-noun phrases. This is important when considering your SEO page markup and page copy strategy. Since verbs, for example account for only 2.4% of all search terms, you would not sweat bullets over their usage in your website. The linguistic profile for human search term usage is:
- Proper noun: 40.2%
- Noun: 30.9%
- Adjectives: 7.1%
- URI: 5.9%
- Preposition: 3.7%
- Garbage strings: 2.5%
- Verb: 2.4%
- Punctuation: 1.4%
- All other parts of speech: 5.9%
So, how does the “home improvement” data set look when considering the importance of adjectives, nouns and phrase word order? There are three things that can be observed about these consumer queries:
First, 66% of the queries have this noun-noun pattern: home improvement [noun]
Second, 23% of queries have a proper noun-noun pattern: [company name] home improvement
The remaining 11% of the queries have an adjective-noun pattern: [adjective] home improvement
There are a few examples that don’t fit this model—but, as a rule-of-thumb linguistic search behavior follows these three simple patterns that we see with the “home improvement” data set.
A search behavior model
The primary thrust of user searches in the home improvement category is for information about products and information about remodeling or repairing a part of the home. Searchers are also looking for ideas, advice and reviews—and are interested in doing some of the work themselves. Quality and affordability are very important, and is reflected in the search behavior.
The secondary focus is searching for companies that provide contracting services, and companies that sell hardware. And finally, there is an interest in financing these renovations.
Modeling a content strategy
Once a search behavior model is understood you can use this information to develop a content strategy. The following strategic content model reflects human behavior in the AdWords dataset. Many users tend to start their search sessions using ambiguous queries such as home improvement remodeling, or home improvement projects. They then move to specific keyword terms, using phrases such as home improvement contractors, Home Improvement Bathroom or home improvement DIY.
The use of the terms remodeling or repairing is practically the same in intended outcome—the searcher wants to improve a portion of his house. Customers are using two different phrases to frame a single concept. From an information architecture perspective you could use these two categories as funnels on custom landing pages depending upon what preferred phraseology is being used.
When you view all the categories from an organizational perspective they fall into the following groups: projects, information, products and companies. This dictates a very simple home page organizational strategy, with four major modules with projects and products playing the most dominant roles, with information and company directory in supporting roles. The following information architecture reflects user intent, and provides a one-stop strategy to provide users with what they are looking for.
Web page copy
When you strip away the primary keyword phrase (home improvement) and look at the top secondary terms by volume, you are left with a set of terms that suggest major SEO opportunities, with a very focused list of words that should be worked into the website’s ad copy.
The following list also provides you with a set of terms that people are using to find information about the Home Improvement TV show (red font). Most of these you do not have to worry about—but, if you had a “series of How To videos and DVD” for sale on your website, you may want to think about providing some “goodwill” value by providing links to the content these folks are interested in. Or you could just ignore this traffic (unless you are running an AdWords campaign, in which case you should consider including these as negative match keywords). These secondary terms, with the number of queries for each, includes:
- tools: 1,000,000
- bathroom: 550,000
- hardware: 450,000
- quality: 246,000
- flooring: 165,000
- plumbing: 165,000
- kitchen: 110,000
- loews: 81,700
- store: 73,600
- tv: 37,200
- set: 33,100
- tile: 33,100
- warehouse: 33,100
- contractor: 32,900
- construction: 27,100
- product: 20,500
- center: 19,200
- remodel: 18,820
- tim: 18,500
- american: 18,100
- repair: 18,100
- cast: 14,800
- companies: 13,500
- service: 13,500
- discount: 12,100
- other: 12,100
- best: 11,800
- dvd: 11,500
- complete: 9,900
- ideas: 9,900
- inc: 9,900
- online: 9,900
- series: 9,900
- wilson: 9,900
- design: 8,100
- diy: 8,100
- projects: 8,100
- video: 7,000
- appliances: 6,600
- city: 6,600
- guide: 6,600
- review: 6,500
- reviews: 6,500
- advice: 5,400
- green: 5,400
- how to: 5,400
- professional: 5,300
- websites: 5,190
- season 3: 4,500
- classic: 4,400
In this analysis you see that a complex set of reported data from AdWords searches around a single topic can be reduced to nine search behavior categories. In my experience, I’ve seen as few as six categories in some travel-related AdWords data sets, and as many as twelve in medical areas.
Resolving the keyword phrases for the top secondary terms provides your information architects with a focused short list of terms that provide SEO opportunities, and content module labeling opportunities. This list of terms also provides you with a preferred vocabulary for website ad copy.
This analysis of human search behavior provides a data-driven approach to developing and refining a content strategy that aligns your information architecture with user intent. If you give the users what they are searching for—they will find you.
Opinions expressed in this article are those of the guest author and not necessarily Search Engine Land. Staff authors are listed here.