£878,111 spent £3.8million in clinical negligence work ‘accepted’ 13,531 medical negligence leads delivered 431 …EXPLORE
How to build buyer personas for FREE using 3rd party dataBy John King | 4th July 2019 Content Marketing
Today, we’re going to use free 3rd party data, and data that you already own to build detailed buyer personas. These will often set you back £1,000 – £3,000+ if handled by a marketing agency, and we’re going to do it in less than half a day using:
- Google Analytics – for location, age and broad interest data
- Facebook Insights – for comparative age and location data, as well as deeper interest data
- Quantcast Measure – more comparative demographic and preference data
- Answer the Public – query insights to highlight what people are looking for
- Quora – to discover the ‘best’ answers to user questions and cues on messaging
Google Analytics and Quantcast Measure depend on you having a website, and having the corresponding tracking and analysis tags installed, as the tags are required to gather the data we’ll be needing. For Facebook, it automatically pulls data from the people connected to your Facebook page, so if you have a Facebook page with enough likes, you’re golden.
If you don’t have the Google and Quantcast tags already, I highly recommend that you get them, provided site-speed isn’t a problem for you. I also advocate getting the Facebook Pixel installed too, but we won’t need that for the purpose of this guide.
For more information on installing these tags, see the following help pages:
Why buyer personas are important
Every good marketing strategy and subsequent marketing plan depends on a well-rounded body of research.
If you know what your ideal client looks like, what interests them, and how they behave, you can make better decisions when targeting your advertising, choosing channels and developing your message and offer.
Sadly, we don’t all have tens of thousands to spend on ‘men in white coats’ to interrogate and experiment with our past and present clients. This is even more difficult if you’re just starting out and you don’t have a great deal of client data to work with – or clients to interview.
For the sake of generating leads via digital channels and social media, you don’t need to spend hours in excel slicing and dicing research figures or blending government and public statistics.
“Heresy!” I hear you exclaim.
If you work in a law firm (as I did at the time) or any other ‘thrifty’ organisation, you may be tempted to develop a comprehensive business case around amazingly detailed research.
Heck, you might need to do this because your finance team needs some convincing to give you the budget.
The trouble is, this kind of desk research rarely translates into real-world performance, and even if it does, all this data is generated for you automatically when you run a test campaign; so why not save yourself a job?
After all, no marketing plan survives first-contact with the end user.
Through much trial and error, I found that the best balance is to get a rough picture of the client to give your early campaigns some structure; after your test campaigns are completed, you’ll use this new data to refine and iterate on your approach which will render any pre-campaign research largely redundant.
In the examples you’ll see below, I’m building a buyer persona to support the launch of a Facebook lead generation campaign which will deliver clinical negligence enquiries. I want to know what these users look like, how old they are, where they’re located, what their pains and challenges are, and how they react to new information to help shape my messaging.
As I go along, I’ll get cataloguing my findings in a table, which we’ll review together at the end of each of the five data sources, feel free to build your own in excel.
Some of the images used throughout will be hard to see and read, if you click them, you can see an enlarged version.
So without further ado, let’s begin.
If your business has a website, there’s a good chance it’s running Google Analytics (GA) or a similar measurement tool. If you’re a massive business, you may even have Google 360 or Adobe Analytics (you lucky swine). If you’re using Google Analytics you can unearth basic demographic insights, very broad interests, and some data on the technology your clients and prospects are using.
Getting age and gender
If you’re a GA user, hop into your analytics account, and on the left, select ‘audience’ > ‘demographics’ > ‘overview’.
You should see something like this:
Here, we see a slight female bias, we can also see that the 18-24 bracket isn’t at all attractive, but surprisingly, the 25-34 bracket is much larger than I might have expected, with healthy traffic in older age brackets too.
You can get more accurate insights by using ‘segments’, where you define a view that singles out particular website pages, categories and/or people based on device usage, etc. We won’t get into setting up segments in this post, but you can find out more on segments directly from google.
In the screenshot above, I was only looking at traffic to pages in the ‘medical negligence’ section of my then employer’s webite – this is important; as other businesses and ‘non-clients’ are likely to check out your homepage and contact page, and you don’t want these visits skewing the numbers you see above.
Think of all the paralegals visiting your site to get your contact details to chase up cases and correspondance – you don’t want them to count.
Getting broad user interests
Now, head to ‘audience’ > ‘interests’ > ‘overview’.
Here you can see the broad topics that people at the top of your funnel (affinity categories) are interested in, and the interests of those users who are likely to convert soon (in-market segments).
In my case, we’re seeing a lot of people interested in travel, motors, real estate, financial services and ‘home & garden’.
Based on what you’re seeing in your own report, is there any way you can blend your product or service with the topics you see in this report? Could you use these insights to tailor your content, adverts or future research?
Understanding user location
We can also see where many of our users are coming from. On the left-hand navigation, go to:
‘Audience’ > ‘Geo’ > ‘Location’ and then in the tiny bar above the button labelled ‘Secondary dimension’, click ‘City’:
Above, we can see we get a great deal of interest from London, which we’d expect given its population density; then we have a sizable following from Birmingham, Manchester and Liverpool. Most law firms in England don’t handle Scottish cases, so sadly, we’d be unable to help those users visiting from Glasgow.
User technology preferences
Finally, we can look at the technology our users are… using. You can go into a lot of detail if you choose to, but for the sake of this guide, we only really need to know what devices our prospects use to engage with us.
To get this report, head to ‘audience’ > ‘mobile’ > ‘overview’.
In this case, we can see that ‘desktop’ visits account for a mere 25% of all traffic, with mobile phones taking the lion’s share at 60%.
- If you see a similar pattern, how well does your website perform on mobile?
- When creating digital advertising campaigns, will this help inform your decision to target ‘mobile’ (phone and tablet) users, desktop users, or both?
Would you be tempted to focus on mobile traffic? Or would you be tempted to run separate campaigns for each?
Google Analytics research summary
After our first research stop we have some insight into client on their age, gender, location, broad interests and device behaviours – which I’ve summarised in the table below.
We’ll add rows and columns as we develop this ‘persona’:
Facebook Audience Insights
This is fantastic if you already have a well-liked Facebook page, or are keenly aware of what brands and activities your audience are into. If you don’t, it’s worth rooting around in anyway as you may still find interesting insights based on the broader demographic picture of your audience.
Before mid-2018, this was uber powerful. You could create custom audiences in your Facebook account, and then use the Insights tool to see what your users liked, their demographics, education, the lot. So, if someone visited a specific product page on your site, assuming there was enough traffic to analyse, it could give you a detailed picture of their ‘persona’
Naturally, someone found a way to spoil it for the rest of us. So now, you can use it to explore the Facebook community more widely, or just those people connected to your Facebook page.
Head to the Facebook Insights tool.
Go into your advertising account > Business Manager > All Tools > Audience Insights
You’ll often see a screen giving you the option to examine ‘everyone on Facebook’ or ‘people connected to your page’ – opt for the latter if you have a page, then change the location to ‘United Kingdom’
Facebook user ages and gender
Let’s have a look at a page I made earlier:
The light grey (or are they light blue?) columns represent the Facebook community at large. The Dark blue columns represent the people connected to my page.
53% of all people on Facebook are women, but of those connected to my page, 73% are all women.
Similarly, we can see greater interest coming from the middle-age brackets at 35+, which is in line with what we saw in Google Analytics
Facebook user interests
Now, what about interests and page likes?
The insights gleaned above can help you to round out your marketing personas, which in turn can help you craft content that’s more likely to resonate with your users. Similarly, many of the mentioned entities are businesses, so can you target their websites with display ads or speak to them about joint partnerships and offers? The use of this insight is limited only by your creativity.
Facebook user locations
Further, you can look to see which locations are best represented by the people who like your page. Be mindful that any historic advertising, and locations with remarkable population density are likely to skew the results here.
In our report we can see a significant number of people from the North West of England taking an interest in our page which echoes our findings in Google Analytics:
Facebook user behaviour and activity
And finally, you can get a view on how they behave day-to-day on the Facebook platform. For those of you building multi-channel strategies, it can be useful to know what devices your audience uses to access the web, and how they respond to different kinds of content:
Above we can see my audience are avid page-likers and respond incredibly well to Facebook ads (it’s worth noting that most of these people will have been driven to my Facebook page by Facebook ads, so this may be skewed by marketing activity) – even so, they are keen commenters, so posts that are conversation starters may work well in your social strategy, and they’re good at increasing your reach by liking and sharing page posts. Finally, their almost exclusive use of mobile devices to access the Facebook platform tells me that I should probably dispense with any campaigns that include ‘desktop targeting’.
Facebook research summary
So, after our second stop, we now have:
As you can see, the data sources seem to broadly agree with one another, with Facebook providing some additional layers of insight and more detail on interests.
So far, so good.
Quantcast is an audience intelligence and advertising platform that tracks users and their interests and uses this data to deliver more targeted ads to them.
This works really well for our purposes, because if you’ve had the Quantcast Measure tag installed on your site for any length of time, you can get insights on your audience, free of charge, without needing to spend a penny on advertising.
Most people I speak to in the law space don’t have this tag installed, but don’t worry if you don’t – if you sign up for an account and get the Quantcast tag installed, you can search for other company and publisher websites to see similar reports. As such, I’ll show you what to expect once your tag has some data to show you.
Under ‘audiences’ you’ll be presented with a list of all the audiences you’ve set up. Ideally, you’ll create these based on the products or services you provide, so in the case of a personal injury law firm, you might create an audience for all people that visit a URL that contains ‘/medical-negligence/’ or ‘/road-traffic-accidents/’ – you can even create something more granular if you like, or even audiences that are site-wide.
In my case, I’m sticking to medical negligence. Here’s what I got:
Notice that groups are measured by ‘index’?
100 is considered the baseline for the population at large, so under ‘gender’ we can see that ‘Female’ users over-index at 128, so considerably more of our users are female, compared to the wider population. Likewise, we can see a corresponding drop in the number of ‘Male’ users, who under-index at 73.
Again, a considerable female bias.
We can also see the 25-44 year old segments over represented here, and that a great deal of our users have children.
Household incomes are broadly in line with the rest of the population with a slight lean towards lower income brackets at £0-30k
Education is conflicting here where the ‘grad school’ equivalent in the UK is postgraduate study. Our sample tells us many of our users are post graduates, but our Facebook sample tells a different story. Given the differences between our audience and the rest of the population aren’t large enough to alter our future marketing approach, we might as well leave this ‘as is’ for now.
Further down the list of reports, we find the ‘cross-platform’ report which demonstrates the same widening gap between desktop and mobile users, with mobile devices making up 65% of traffic, while desktop makes up just 35% – again, a good reason to certain that our user experience plays out well across mobile devices:
We can also get some insights into the location of our users:
Here, we can see interest from Birmingham, the North West (Manchester & Liverpool) and London at the bottom. In this sample, London doesn’t skew the results as heavily as it did in our Google Analytics and Facebook Insights reports – but broadly they all agree with each other.
The Your Legal Friend website at the time, had a Polish version too, which explains the interest from Warsaw and Krakow and the other Polish countries listed.
Depending on the amount of traffic your site generates, you’ll see reports with differening levels of detail and accuracy. For this property, there isn’t enough data to see ‘browsing interests’, but we can see some basic ‘Shopping interests’ split among a multitude of categories.
In this case, our users are more than twice as likely (index 222) to be Vauxhall fans:
They’re more likely than most to splurge on consumer electronics:
They’re twice as likely to be theatre and musical lovers:
And then we can see their spending patterns:
There are loads of others we could look at, but I’ll leave you to explore those in your own time. Instead, I’ll jump to the ones that might help you with ‘buyer persona’ building.
Under hobbies, we see that some of our audience over index against ‘photography’:
They’re big fans of gardening and housewares, which ties in nicely to our findings in Google Analytics:
Further, our visitors are big fans of blogs and social networking, while spending more time on the Internet than most – this is particularly helpful if you need to help justify any investment in content marketing, digital or social media advertising:
And for those that like the outdoors, we can see that fishing, hiking and water sports are a big draw:
Moving on to ‘Psychographics & Lifestyles’, we see a considerable interest in health and wellness, as well as going to the pub:
And to take this further still, we can see that Quantcast has already compiled some of our users into persona categories; where fashionistas, gadget enthusiasts, mainstream buyers and football enthusiasts make up a sizable proportion of our audience:
We can even see some broad patterns in preferred TV genres:
Once again, you can scout out other websites that use Quantcast to see what their audiences are interested in. The reports for the lawgazette.co.uk show heavy preferences for luxury cars and are three times more likely than most to be earning over £70k; so, if the stereotypes are true, the data Quantcast provides must be fairly accurate:
If you think back to the interests we saw in Google Analytics and Facebook insights, you could look up one of the websites mentioned to see if you can make use of the audience data for that site instead.
Quantcast research summary
Taking together all of our research so far, we can see that with the exception of education, our data sources broadly agree with one another:
Not bad for half a day spent browsing free data sources.
Now, we need to connect the customer to the product or service we’re marketing.
For that, we’re going to use Answer the Public and Quora.
Bear in mind that we’re making a sizable leap in assuming that the prospects we’ve researched will be interested in what we have to offer. If your findings, like mine, are based on data from users who have visited your site with come conversions and enquiries taking place, then this is less of a risk. If you’re offering an entirely new product or service, just be mindful that more research might be needed.
Let’s start with Answer the Public.
Answer the Public
In my case, I’m looking to run a campaign targeting people who feel that they’ve been mistreated by medical professionals. For those who are familiar with the terminology, medical mistreatment that leads to provable injury is often referred to as ‘medical negligence’ – so let’s start by searching for that:
Assuming the engine finds enough query data to report on, we’ll be presented with questions and themes around the topic we’ve entered in diagram format, like the one below:
While hard to read, of note in the image above are queries like:
- medical negligence how long after can i claim
- medical negligence can i claim
- is medical negligence a criminal offence
- medical negligence for a child
- medical negligence for cerebral palsy
- medical negligence for endometriosis
- medical negligence solicitors near me
- medical negligence lawyers near me
If you’re following along, you’ll see there are other images and graphs, but the next one we’ll look at is the ‘related searches’ wheel:
Above, queries worth our attention include:
- medical negligence examples
- medical negligence cases 2018
- medical negligence payout
- medical negligence solicitors
- medical negligence nhs
- medical negligence claims
- medical negligence experts
- medical negligence solicitors manchester
- medical negligence calculator
- medical negligence lawyers
- medical negligence compensation
- medical negligence time limit
Most users who are searching for ‘medical negligence’ are seeking legal help, or information around the subject that lends itself well to a ‘legal’ viewpoint.
The content marketers among you will already see topics here that could be expanded on and scheduled into a content calendar, while those of you with an interest in SEO can see the contribution this sort of insight can make to your keyword research.
Broadly, we can deduce that people searching around the topic of ‘medical negligence’ are interested in:
- Examples of what ‘is’ medical negligence
- Publicised cases by year
- Possible legal solutions
- How to make a claim
- Whether they can claim in the first place
- The time limits on making such claims
- Possible payout amounts
- Medical negligence incidents involving children
- Incidents involving Endometriosis and Cerebral Palsy
- Medical negligence within the NHS
- Whether there are medical negligence solicitors close to them as well as location specific searches (Manchester)
Answer the Public research summary
There’s no need to follow my table structure above, you can collate this information however you see fit. From the insights forming above, it’s interesting to see the niche specific searches around childbirth and medical negligence, when you consider our sizeable female bias.
Could this be the reason why we’re attracting so many women and families with children in the first place? Or have we simply spotted a way to better inform our female heavy audience about the dangers of medical negligence in specific regard to childbirth and children?
We can see that some searchers are indeed interested in our service and are asking questions that we’re well positioned to answer.
We can now add further insight to these findings by exploring precisely how other people ask these questions, and what discussions provide the most desirable answers.
For that, let’s head to Quora.
Using Quora to understand user intent
You’ll need to sign up for a free account, and you’ll need to pick 10 topics of interest to continue to the main dashboard. Once done, you’ll see a stream of questions and discussion related to the topics you chose.
At the top, enter your desired search term:
You’ll then see a search results page like this:
Note how these questions already share similarities with the query analysis we saw from Answer the Public? Let’s select the second result as it’s closer to what we offer.
In doing so, we see the question and the most ‘upvoted’ answer, considered to be the best for this particular question.
In case above, there are 17 different answers to choose from, but most converge along similar lines – citing the importance of having experts involved to handle the case, and precisely what it is that constitutes medical negligence – I won’t bore you by screenshotting them all here, but another, very detailed (but American) answer stands out:
The answer is much longer than what you can see here, but the detailed account of how the process works appears to also be of interest.
People therefore aren’t just concerned with knowing the answers, some are also keen to understand how the process works and why it works that way.
Building on the queries we discovered through answer the public, we can search for more specific instances.
Let’s explore the link between medical negligence and children:
As we can see from just these two examples, the answers and discussion are wide and varied.
The top question is closer to the sorts of incidents we would expect in our field. When looking at the answer and subsequent discussion, we can see a clear attempt to provide professional insight and balance – indeed, the ‘top’ answer was upvoted more than 3,000 times:
Further legal perspectives from this question follow:
Again, many of these discussions are American, but appear just as readily for searchers from the UK.
Quora research summary
If you spend more than ten minutes reading through the discussion and the responses, it’s clear that the topic is sensitive and emotionally charged.
Further, there’s a lot of love for the ‘tired’ surgeons and doctors that deliver babies.
This has implications for your adverts and your writing.
In my case, it’s important that any content produced to ‘answer’ these questions is written in a balanced way that shows appreciation for the hard work that doctors do. At the same time, the language must be sensitive and personal. Further, there’s clearly a hunger for an understanding of the medical and legal complexities involved, so expect to write adverts and content that speak to this need to better understand the topic.
Let’s add these findings to our table:
Taken together, our findings paint a detailed and engaging picture of our target audience that will help us to shape the messaging in our adverts, our interest-based targeting as well as the channels and platforms we may choose to use to engage with our broad buyer persona.
We can also take lessons from what we’ve seen to inform our keyword research and SEO strategy, as well as providing ideas for future content marketing and blogging efforts.
If you can, take the time to sense-check your buyer personas with someone else in your business who speaks to customers and clients regularly. They may be able to add insight to your findings to help round out your personas, and they may see patterns that they can explain that desk research simply can’t.
If you’ve made it this far, I hope you found the guide useful!
Drop me a message with any comments or feedback.
If you’ve spotted any additional data sources, trick or techniques that you’ve added to this approach; let me know! I’d love to reach out to you for comment, and some examples to update this guide with.