As digital marketers enter their second year in the post-Hummingbird world, it is becoming more and more obvious that keyword-level research is too shallow to be relied upon for consistent, credible results. Market segmentation, on the other hand, enables marketers to gain a deep understanding of their audience and break that audience down into manageable, predictable groups.
Market Segmentation: The Art of Categorization
By definition, market segmentation is: The aggregating of prospective buyers into groups (segments) that have common needs and will respond similarly to a marketing action. Market segmentation enables companies to target different categories of consumers who perceive the full value of certain products and services differently from one another. Generally, three criteria can be used to identify different market segments:
- Homogeneity (common needs within segment)
- Distinction (unique from other groups)
- Reaction (similar response to market)
Market segmentation breaks large target audiences down into smaller, more specified groups.
Basically, market segmentation is a strategy that splinters a broad target market into subsets — which are based on common needs and interests — and then helps marketers develop a plan to target those subsets. For example, the broad target market of men who buy sneakers could be broken down into the subsets of men who buy sneakers for fashion and men who buy sneakers for sports.
Persona-Driven Market Segmentation
A big group of buyers becomes many little groups of buyers.
One strategy uses fictional characters to represent those little groups of buyers. Those fictional characters are called audience personas. Creating reliable and accurate audience personas is the most crucial aspect of successful market segmentation.
Digital marketers are working to adopt the strategy of persona-driven market segmentation because:
- Google is clearly pushing all channels in this direction, not just through the release of Hummingbird, but also through the integration of Google Plus as an “identity platform” as well as a big data Affinity Segments platform.
- Conversion rates improve when audiences are identified and directly targeted.
- Understanding the subtleties and nuances of your audiences helps you weed out bad leads and increase qualified leads for your sales team.
- Maximize your traffic and make your site stickier by drawing more of the people who are looking for what you have.
Developing Personas: The Soccer Mom Phenomenon
In the weeks leading up to the 1996 presidential election, Republican strategists coined the term “soccer moms“. Their candidate, Bob Dole, was eager to appeal to a group of voters who — if neglected — could swing the election in favor of his opponent, Democrat Bill Clinton.
This vast group of still-undecided voters were white, married, middle-class, suburban mothers who owned SUVs to shuttle their multiple children to the their after-school athletic activities, and then back to the homes they owned in housing developments.
They were, of course, individuals in different parts of the country with different stories, different lives, different beliefs and different life experiences. But through research and analysis of massive amounts of data, pollsters separated them from the larger population and grouped them into a subset — a persona.
Bob Dole lost the election, but a powerful new voting block had been accurately identified — soccer moms were here to stay.
Just as political scientists isolated a group of the national electorate and predicted how they would vote, persona-driven market segmentation helps marketers isolate and identify potential customers and how they will buy.
Rapper-turned-web-developer-turned-SEO-turned-digital-marketing-consultant Michael King compared the keyword model to the persona model this way: Keywords are stick-figure drawings and personas are action figures. Both are abstractions, but personas are three dimensional, more robust and more realistic.
Buyers vs. Audience: Understanding the Different Types of Personas
The many different kinds of personas can basically be grouped into one of two categories: buyer personas and audience personas.
Audience personas are on your site to consume content, not to buy. They want to read your blog or scan your articles for educational or entertainment purposes, and although you’ve successfully attracted them with ads or SEO or search keywords, they aren’t going to buy anything.
Buyer personas are exactly what they sound like. They’ve come to your site because they’re shopping for what you have. They, too, may have visited to consume content — but they’re doing it as part of a preparation to make a purchase.
It is important to note that there is often crossover between the two personas, but even though the two groups often overlap, treat them as independent entities.
Building Personas Part 1: Google Affinity Segments
In June 2013, Google launched Affinity Segments, which uses big data to model people based on enormous data sets that the ubiquitous search giant perpetually collects. It is complicated, and largely for that reason, marketers have generally not embraced it en masse.
But with or without them, Google is moving in the direction of audience-based customer targeting.
In Google’s own words, “adding audiences allows you to reach people based on their specific interests as they browse pages, videos, and content across YouTube and the Google Display Network as well as channels and videos on the YouTube Search Network. You can select from a wide range of categories — from autos and sports to travel and fashion — and we’ll show ads to people who are likely to be interested.”
Depending on your goals and the stage of the purchase process your customers are in, you can add different audiences to your ad groups and target groups, including:
- Affinity Audiences: These audiences were built for businesses currently running a TV ad who would like to extend the reach of a TV campaign to an online context for an efficient price.
- Custom Affinity Audiences: With custom affinity audiences, advertisers can create audiences that are more tailored to their brands, compared to our broad, TV-like affinity audiences. For example, rather than reaching Sports Fans, a running shoe company may want to reach Avid Marathon Runners instead.
- In-Market Audiences: These audiences are designed for advertisers focused on getting conversions from customers most likely to make a purchase. In-market audiences can help drive remarketing performance and reach consumers close to completing a purchase.
- Other Interests: Use these additional audiences to reach customers who may be likely to visit your site. You can also use these audiences to show your ads to people who have interests that aren’t included in the affinity audiences or in-market audiences.
How Audiences Work
People use incredibly powerful and helpful technology like Google and Facebook for free — sort of.
They pay with their data.
Google has both the sample size and the amount of data per sample to produce results that are more robust than that provided by any other entity. So how do they do it? According to the Affinity Segments page, “When someone visits a Google partner website, we may use the content of the page or website to show contextually relevant ads. We use the main topics and themes from the page as well as data from third-party companies to associate interests with a visitor’s anonymous cookie ID. This takes into account how often people visit sites like those, among other factors.
Google may use information that people provide to these partner websites about their gender, age, and other demographic or interest information. We may also use the websites people visit and third-party data to infer this information. For example, if the sites a person visits have a majority of female visitors (based on aggregated survey data on site visitation), we may associate the person’s cookie with the female demographic category.”
Building Personas Part 2: Google Analytics and Social Media
Remember the difference between buyer personas and audience personas? Audience personas visit to browse. Buyer personas visit to buy. You can go beyond ads by using Google Analytics to create buyer personas for your social-media campaign, as described in one expert tutorial.
The goal is to learn about your online audience in the search data from your website’s analytics.
Step 1: Investigate Your Traffic by Keyword
Go to Google Analytics -> Acquisition -> All Traffic -> Google/Organic and then set the Secondary Dimension to “Keyword”. This gives you a long history of the keywords that brought people to your website.
This is a woefully incomplete list, but it’s a good jumping-off point.
Step 2: Scour Search Traffic for User Similarities
Now that you’ve got your keywords isolated, group them into themes. If you sell electronics, for example, create groups such as “audio”, “televisions” and “HDMI cables”.
Now brainstorm about who is searching for these items. Is it likely to be DJs and other professionals, media consumers or a mix of different groups?
Step 3: Break Down Personas by Social Channels
Now create archetype audiences for all of your social-media platforms. Accomplish this by using the referral traffic data from Analytics. Go to Google Analytics -> Acquisition -> All Referrals. Choose Second Dimension and click “Landing Page”.
Isolate this data and group the landing pages by social channel to find out which content works the best with each social-media network’s users.
If you own a bike shop, more people looking for repairs may come from Twitter, where Facebook may refer more people looking to purchase new bikes. Now you know to pitch your repair services on Twitter and your new sales on Facebook.
Building Personas Part 3: Qualitative Research
There are many ways to build personas. Michael King’s model is based on the premise that narrowly focused research based on too few sources will put too much emphasis on your own assumptions — and that isn’t research at all.
A blend of qualitative and quantitative research can ensure marketers are getting the big picture when building personas.
According to King, “with qualitative research, you’re asking open-ended questions to small sample sizes to get a sense of the hows and the whys behind a specific problem. You’re typically looking at unstructured data to inform commonality amongst your user group and any insights are then validated for scale throughout quantitative research processes. Qualitative research within our context is often user interviews, focus groups, content analysis, text-mining, ethnography and affinity mapping”.
When it comes to Qualitative Research, King suggests:
Affinity Mapping/Affinity Diagramming
Affinity mapping (or diagramming) is the process of using post-it notes to collect and organize everyone’s thoughts, visualize and discuss them as a group. The exercise entails three, 90-minute rounds. Assumption Round 1 addresses needs, Assumption Round 2 addresses attributes and the Factoid Round involves gathering facts to back up the assumptions.
Finally, whittle the information you’ve collected down to three or four “skeletal personas.” Can’t get everyone in the same room because of geographical limitations? Use Mural.ly and Google+ Hangouts to affinity map remotely.
Gather people in the target segment into a moderated meeting. The moderator asks questions about the needs, desires and biases of the group. Focus groups can produce useful, real-world insight, but it’s important to have an experienced moderator — and be warned that it only takes one strong-headed participant to steer the whole group.
Similar to focus groups in design and intent, customer interviews provide the intimacy of a one-on-one discussion, without any external influences.
Building Personas Part 4: Quantitative Research
According to King, “Quantitative research is about using numbers and statistics to understand behaviors of users empirically. The sample sizes are often quite large so that the insights can be applied to broad populations of people.”
He recommends basing quantitative research on:
Multiple-choice questionnaires are built for precision. They should focus on specific questions that are NOT open ended. Offer well-defined choices that are based on your qualitative research.
Ready-made tools such as MRI and Comscore help to build personas by offer pre-fab segments. Be warned — these tools require a lot of data input and a wealth of information to work properly. They are not an overnight fix — be patient!
Analytics, User Profiles and Internal Data
Analytics, of course, depends on your setup. But look at internal search, paid search and historical organic search keywords. Does your site allow user profiles? These are incredibly productive places to look for data, especially if they use data from social logins. Internal data may include sales calls, returns and reviews.
Finally, leverage knowledge gleaned from publicly available studies. Check out Google’s Consumer Barometer for public data points.
Become the Persona By Asking the Right Questions
Consumer personas give businesses a way to see how the things they’re doing are being viewed from the outside. In order for it to work, you have to think like the persona does. The problem is, the persona doesn’t exist.
One expert compared persona-driven market segmentation to method acting, where an actor “gets into character” by immersing themselves in the person they’re attempting to emulate.
In order to get into character, it is crucial to ask lots of questions of your persona when attempting to think the way they think — or would think if they were real.
Big and Obvious — But Necessary — Questions
Demographic information is obvious. Are they married? How much money do they make? Do they live in a city, the suburbs or a rural area?
Once you get past big-subset questions, be sure to determine what they’re likely to do for a living and — this one is important — their level of seniority at their job.
More Abstract Questions Reveal Important Nuances
From there, it gets more subtle. Try to picture a day in their life. What do they do when they’re not doing what they do for a living? Where would they rather be when they’re not where they have to be? How much TV do they watch? How often do they buy new clothes? The author recommends at this point actually going through some stock photos to get an image that literally looks like the person you think your persona is becoming.
Next, find what the other calls “pain points”. What worries them? What don’t they have?
Finally, find out where they get their information, what they value and — most importantly — what experience they’re looking for when they shop.
Your persona may be a kid named Beth, a man named Edgar, a mom named Melissa or a Coach named Brian. But remember, your persona is real — it has to be if this is going to work.
There is no right way to build an audience persona. There are many wrong ways. The wrongest way seems to be when marketers substitute their own presumptions for actual data. The more information gathered, the better. Stand on the shoulders of Google — they have access to more data on the people you’re targeting than what you could ever compile on your. Use their resources, but make sure to supplement it with your own internal research. Facebook Ads can also be an incredible source of data.
Remember, you’re not creating a spreadsheet or a demographic. You’re creating a person. Become the persona, understand the persona and, eventually, you will sell to the people the persona represents.
Note: The opinions expressed in this article are the views of the author, and not necessarily the views of Caphyon, its staff, or its partners.