As marketers, we often have major challenges in understanding our own target audiences, defining those audiences, figuring out the behaviors, and then knowing how to actually make use of those insights.
That can be a little elusive, but that’s exactly what Jonathann Mingoia is going to talk to us about in this episode of The MarTech Show, hosted by Agorapulse’s chief storyteller, Mike Allton, and co-host Robin Dimond. Listen to the whole podcast below or read on for the edited-for-length transcript.
Jonathann is a renowned audience intelligence specialist with over a decade of experience in the field, currently serving as the chief enterprise solutions officer at Audiense. With a robust background and utilizing data-driven approaches to understand consumer behavior, Jonathann’s helped businesses enhance their marketing strategies, optimize their operations, and achieve significant growth.
Tell us little bit more about Audiense and your role there
Jonathann Mingoia: Audiense basically provides audience intelligence through SaaS platforms to brands and agencies. Today, my role at Audiense is mainly focusing on two areas: The first area is in bridging our expertise coming from our solution tool with our solution team—with the data capabilities we have from our different tools to achieve more advanced enterprise use cases—but also working alongside the success team we have in order to monitor the value alignment of our customers and also supporting our customer to achieve their use case and elevating the full value they can get from our tool.
Robin Dimond: Love your background in social media and where you came from, and thank you for helping us understand Audiense. I think for everyone, and the question is (if you could help us with this):
How does audience intelligence differ from the social and digital intelligence that you’re talking about?
Jonathann Mingoia: Let me start by demystifying what is audience intelligence by highlighting what audience intelligence is not.
Audience intelligence, to avoid any confusing association with a different solution existing, uses social data.
Audience intelligence is not social listening. So we are not providing monitoring of conversation and tracking how it’s evolving, the tonality of topics and conversations across different social channels. But we are super-complimentary to social listening because we have the ability to describe who are the audiences who are generating those conversations online.
Audience intelligence is not social analytics, so we are not providing metrics about engagement or ad performance or any metrics related to the activity of community margin. But we can complement that set of insights by providing a full description and understanding about, for instance, who are the audiences that are active behind a Facebook fan page or an Instagram account.
“Basically, audience intelligence is audience understanding by using a combination of multiple methodologies and data points to develop personal segmentation, understanding the passion point of audiences, and also getting a sense of what is the value of the audience.” (Jonathan Mingoia)
This is a crucial layer before starting any kind of operation in marketing or communication. So when we are talking about, “Who is my target group? What are the different segments existing in my audience? And what are the unique patterns of each of those different segments? What is the affinity with brands? What type of influencers can create an acquisition of net new audiences? Or, for instance, what are the personality traits of the different segments that are composing my audience?”
- We are talking about use cases that are fully covered by audience intelligence. And, yes, typically, the methodology behind audience intelligence is collecting information from different social channels through API in a compliant way—with no scraping data in a way that does not respect the terms and conditions of platforms.
- We’re gathering this data, structuring this data, augmenting this data, and enriching these insights by clustering, adding, scoring, cleaning, and aggregating different sets of information to create new categories that do not exist from API—and we are delivering that knowledge about audiences to our customers into self-service platforms, SaaS platforms that they can use autonomously, brand and agencies, and where they will get clean visualization and chart that are highlighting, “What are the key patterns that describe the audience? What are the key brands with who the audience has a strong affinity and all those insights, you know, to improve, and support marketing and communication operations?”
You mentioned what was the difference between audience intelligence and social and digital intelligence.
Audience intelligence is the big umbrella, and in audience intelligence, we are using different methodologies as we have described, and among those methodologies, we have social intelligence and digital intelligence.
Social intelligence is the foundational layer of the audience understanding where, for instance, we are working on an individual level by using proxies such as X to understand the following graph of each member and understand through the interconnectivity of members. What are the unique segments that are composing an audience?
- This is a strong layer of audience intelligence because this is very well connected with social listening since a lot of conversations are coming from it when you are doing a social media monitoring conversation, and when we add the X authors, we can understand who are the people who are generating those conversation.
- And, on top of knowing who they are, we can understand all the interests they have, but they have not mentioned them in their conversations. So it’s an extension of understanding and adding that layer about who is the audience.
- When we are talking about digital intelligence, this is another methodology that works mainly with advertising API insights. (Soprism, for instance, the company that we have funded and that has been acquired by Audiense is working with the Meta advertising API. We have the ability with those insights to work on an aggregated and anonymized level, tapping into 2.8 billion active users worldwide through the members of Instagram, the members of Facebook.)
- On top of having the possibility to describe all the audiences behind the Meta-interest category, we can also—because this is anonymized, aggregated information—generate audience profiling, reporting on first-party audiences (for instance, website visitors, and customers).
- Audience intelligence is the combination of these two types of methodologies. By combining these types of tools that are typically used by our customers alongside each other, we can capture 100 percent of the online audience. It’s the core foundation of any audience intelligence project. [They] complement each other because the individual level enables one to identify the cultural segment of any audience, and the aggregated level enables one to tap into the biggest digital panel existing worldwide based on behavioral data. With that full set of insights, you can get a 360-degree understanding and holistic view of any kind of target group for brands.
Mike Allton: I’m glad you mentioned the API that you use and the compliance factor. I think that’s important for folks to understand when we’re talking about this kind of audience analysis. It is aggregated. It’s not personal data. You are compliant with GDPR and all the other regulations because you’re not identifying that “Mike Allton did this action or took this action.” I’m part of a group segment. So thanks for sharing that.
What other kinds of tools and technologies are essential for this kind of audience intelligence to even take place?
Obviously, you need to have more than just a website. What else do they need?
Jonathann Mingoia: We talked about previously the fact that audience intelligence is based on social and digital intelligence—but for an effective audience intelligence strategy, it’s 100 percent key to diversifying the source of data. We have observed that, for instance, audience strategy success is directly correlated with the diversity of the data source we are using.
It’s quite normal because when you want to understand, for example, the nuance and complexity of consumer behaviors, it comes from combining different lenses of analysis and observation to get that full picture of a target audience.
A recent study from Torch Group media highlighted that only 14 percent of marketers are relying on two or fewer audience inside sources for their audience strategy. Half of the marketers’ respondents in that study are using between three and five different audience insight sources, and 20% are using seven or more data sources when they are building their audience strategy.
This last group those that use more audience insights sources is the group that generates the best business outcome. It is the reason why we consider that the foundation of audience intelligence should combine methodology, and we are providing those types of solutions with audience insights, which is social intelligence and Soprism, which is digital intelligence—also because this type of audience data source is quite powerful because it’s ongoingly updated.
As this is based on data coming from the activities of users on social channels, when you’re generating a study or profiling, you get insights from the moment you have generated that analysis. And if it happens, some shift in the cultural patterns.
The moment you are generating the profiling, you get that timely view about this intelligence, these different insights.
The second point, which is also quite important, is the fact that by combining, social and digital, you can have global coverage with the possibility to get insights on a global level with 3 billion active users on Meta—but also on a national level, on a county level, on a state, on a city level.
It’s representing quite the overall digital population that those sorts of insights.
“The last point that is quite interesting is also: As we are generating audience insights based on behavioral data, you can’t lie. Your real activities and your real consumption of content reflect who you are. And your behavior online cannot be modified or cannot be changed in a way that you want to be perceived in a different way than you truly are.”
It’s important that the fact that behavioral data does not mean biased information.
But on top of the social and digital intelligence solution, other sources of data could complement that stack. For instance, social listening. We talk about that. It enables tracking how the interests, concerns, and feedback of consumers are evolving over their digital channels and also observe and identify the new emerging trends coming from conversations in different communities. For instance, online panels or surveys based on this case in declarative based on declarative data. In that case, it provides a complementary angle of understanding the audience by knowing, for instance, the media consumption and the intent in purchasing a product but also advancing to economic insights.
Another layer is doing deep diving in a small group. So do a focus group where we can have a very small, a very small group and where we can do qualitative analysis about, for instance, “What the feedback they have about packaging or what the motivational triggers to buy a product? What are the main barriers to buying one product versus the competition?”
It’s the last level to get the last details that it’s enabling to get the full understanding of audiences. We know that big corporations are also integrating into their audience intelligence stack a lot of first-party data. So, for instance, point of sales, information, transactional data, and reps browsing behaviors and activities are also integrated to get that full picture and to build that strong layer of audience intelligence in an organization,
Robin Dimond: When we’re looking at this, there’s so, so much to unpack right there.
When we’re looking at this, when we’re looking at the behavior—like, how are behavior analysts on platforms like Meta, how do they decrypt the audience behaviors?
How is Meta qualifying its members?
Jonathann Mingoia: We know that Meta is super powerful, in doing that job.
I think that one of the main values today of Meta is the knowledge they have of 3 billion digital users and they’re gathering and qualifying their members by using different social signaling and different methodologies.
One of the methodologies is when we are creating a profile on Instagram, Meta, Messenger, or WhatsApp, we are sharing some information, but they are also gathering information in real-time from our mobile device or devices where we are going to the Meta platform. They know exactly where we are going out of our homes and where we are going on holidays, so they track and can get information from our devices. This is information based on a deterministic approach.
But the biggest layer and the biggest knowledge that Meta is building is by inferring, interests, and preferences to their members, even if they are not explicitly saying that they have an interest in a brand, lifestyle, sport, or any kind of other passion point and how they are doing that, they are analyzing everything we are doing inside Meta and outside Meta.
In Meta, they are analyzing our visible interaction, like comments, and sharing, if we are doing some publication, some post, for instance. But also in visible interaction—when I’m clicking on a nut, when I’m watching a video post, when I am going into the search engine and looking and searching a specific brand in the sourcing giant of Meta, going to the marketplace, and searching some type of product—all those invisible activities are gathered by Meta to enrich what they know about each of us.
But there is another quite powerful layer is the content consumption approach. When I’m going into my Instagram news feed I’m scrolling to consume my content passively with no interaction. My average feed scroll is analyzed by the algorithm of Meta. Meta can identify that I’m paying attention every time I’m faced with a product coming from a specific brand or any type of content linked to a specific topic. I’m spending more time.
Everything we are doing inside Meta is analyzed in a very precise way. But also when we are going out of Meta, on the website where there is a Meta plugin or pixel of tracking installed in order to do remarketing. Meta is using those pixels of tracking in order to enrich their understanding about what are our preferences and interests. So it’s super powerful. The aggregation of all those social signals enabled us to get that full understanding of any kind of audience.
I remember five years ago, a study led by the University of Cambridge and Stanford showed that from the moment a Facebook member likes more than 150 pieces of content, Meta knows us better than our own family.
It’s super powerful and indeed, I think we can trust in how Meta is inferring through their probabilistic model interests and preferences to each of their members.
Insights Even Meta Won’t Give You
Mike Allton: You talked about Meta a little bit. You talked about using their ad platform and how a lot of the insights are coming from that.
What are some of the insights that you’re going to be able to surface for us that the native analytics provided by platforms like Meta wouldn’t include?
Jonathann Mingoia: Let me show you a concrete example.
One is profiling Facebook’s fan page. When you have the right authorization or the Instagram account, when you have the authorization, or you can also import and profile any kind of custom audiences you can create into the business ad manager of Meta. So based on email address listing, mobile phone numbers, people who are watching some videos on your social channels or website visitors, for instance.
The third level is open audiences. So the Meta interest category. So you can directly, for instance, tap into the 2. 8 billion active users and say, okay, I would like to profile audiences only located in the US. with a specific age range, only active on Instagram, select a gender, a language, and tap into all the Meta Interest categories.
I want to know who are the people interested in the brand Puma, and also, on top of the interest in Puma, interested in sneakers, for instance. Okay, so we have an audience composed of 6. 2 million people based on the different conditions I have selected in my audience, and we are comparing this target audience with a benchmark audience that can be edited.
So in this case, it’s the same sociodemographic group of my target audience, but I couldn’t select Puma versus Nike or Adidas and have the right lens of analysis in order to generate the insights that matter and that enable me to differentiate and position my brand versus the competition.
When I’m launching a profiling, it takes 15 minutes, but what I get has output and insights is, for instance, this is a project Puma Sneakers Lovers in US and what we get typically is yes, it’s an overview where you have the main lifestyle profile that are composing your Puma sneakers lovers in the U.S. So the fashionista, the active sporty, the brand lovers, the interior inspiration seeker, the beauty squad, there are more than 150 lifestyle profiles that are integrated.
And where you have the ability to see exactly, who are those over-indexed in my audience? Having also a story that describes what are the main patterns and affinity of your audience? What are the main sociodemographic elements that are over-indexed with what other brands they have a specific affinity for? What are the devices they are using with what media they have a specific relationship with? What is their relationship with events, personality, and organization? What is the main content that resonates with those Puma sneaker lovers?
And what are the set sociodemographic groups combined with lifestyle dimensions that are key in your audience? And this is only a summary of the key insights, but you have much more by knowing all the sociodemographic elements of your audience in what specific states or cities there is an over-indexing of the Puma sneakers lovers classified, for instance, by affinity or with what order brands, if I want to track among 7, 000 interest with other brands, they have an affinity in the for instance in the fashion industry, I can have a full description of with all the brands they have a specific affinity with.
So I think it was more powerful to highlight and show concretely how you can get insights from a digital intelligence platform, based on the fact that we are profiling an audience and we are gathering as an output 150 lifestyle segment ranking, 500 main topics and also 7, 000 some criteria to check the overlap and affinity with brands, with personalities, with media, with lifestyle interest on the, also the relationship they have with society issues or kind of topic related to potential brand positioning.
So yes, as you can see, it’s quite powerful and, you can get lots of insights that are completely out of the typical basic, sociodemographic insights you can get from an audience. In this case, you can also profile those first-party audiences. Facebook and Instagram account followers and using the benchmark in the good way in order to have the right lens of understanding of that audience that matters for you.
Robin Dimond: When I was looking at it, I got so intrigued by everything and how, how powerful it is for a marketer to use, especially nowadays, especially with everything that’s going on.
Can you give a little bit more about the insights that support like tactical marketing initiatives as you’re showing all of those things?
Jonathann Mingoia: Typically, audience intelligence is the foundation of any kind of marketing and communication operation. It’s really at the center of any initiatives, on the business side for innovation, for new product launching, for promoting, for communicating, and positioning a brand.
But what we are observing is we have three main areas of use cases that are mainly adopted by our customers.
Everything links to product and brand positioning is an important one where brands want to analyze the penetration of their competitors in a specific product category, specific geolocation, in a specific target group.
Second use case: In terms of product and brand positioning is evaluating, “What is the product interest before launching a product? What is the receptivity? What is the size of the potential target group by evaluating the level of interest for an alternative product or the interest, or the product category if this exists already?”
Also, on a more corporate level, “What are the best corporate social responsibility approaches that resonate with the brand value of the brand value, but also the expectations of the target group?”
These are the three main ones for the brand and product positioning.
On the research part, it’s quite powerful and mainly used by marketing managers or marketers in general, which is personal validation. They have a description of personal marketing, and they want to validate if those personas reflect reality in terms of their target group. The people who have a strong appetite for the product or the services that the brand wants to deliver, or refine the persona by enriching or by sifting when there is some cultural evolution, how are shifting and evolving those personas?
A second use case based on the consumer insights part is about generating deep customer profiling. So in that case, they are using their first-party data, CRM email address listing, and they are profiling those email address listings to be able to identify what are the main pattern that describes those high-value clients and how to reach those lookalike audiences not yet acquired, but that have a similar pattern of the customer acquired and about the tactical use case.
Everything links to partnerships—which could be sponsoring events, or brand associations by checking the overlap between the audience in the brand—but also the event audience is quite powerful to select and fine-tune the selection of association and partnership.
Content ideation, personalized content by four different segments is also a tactical use case. Quite important. Identifying in what area of influence makes sense to create a specific relationship with a personality to gain a net new audience by the fact that we are reaching a new community in a specific area. Advertising relevancy boosting, because when you know exactly who is your audience and you know for each of the different segments composing your audience, what message, call to action, and type of content is fitting.
Also, what are the best types of partnerships you can create? In that case, it’s increasing the overall effectiveness and the relevancy of all your marketing initiatives to your target group.
Potential Challenges (And How to Overcome Them)
Mike Allton: Now we’ve got folks listening [and reading!] now and in the future who, let’s say, they’ve gone through your process. They’re starting to do some audience analysis, and they’re concerned they don’t want to make some of the common mistakes.
What are some of those challenges? What are some of those pitfalls that you’ve seen folks run into in the past, and how could they potentially overcome them or bypass them completely?
Jonathann Mingoia: It’s a very, very interesting point.
The first point is about the implementation of a no-gain strategy approach in the evaluation of what could be the best solution to adopt in a company. There is a first challenge, which is the lack of transparency because, in our industry, sometimes they’re not super transparent about the methodology they are using when we are talking about data collection.
You have mentioned that previously scraping data or gathering data in an noncompliant way could be an approach adopted by players.
One of the challenges is evaluating in a good way and removing any black box thing when we are in the evaluation of what the best solutions fit with the brand policy of an organization. It’s very important to identify vendors that are transparent on that side by challenging them about how they are collecting the data, and what methodology they are applying behind the scoring they’re generating. Sometimes you have sample data that are used and you have a scoring, but you don’t know that this is not based on the full data scope and only on the sample. So it’s really important to get full transparency about the methodology and the way the different vendors are collecting the data.
Other Martech Show episode: How to future proof your brand’s marketing
When you have implemented your audience intelligence strategy by selecting the tool that fits your expectations, one of the challenges, it’s about the interoperability of those solutions with other tools Tech stack. We talk about diversity, diversifying all different audiences inside the tool, but if you’re not able to complement and create that link between one audience from one tool to another one, you are potentially creating more complexity and less clarity.
So, it’s key to identify vendors that have in their core DNA a connection with the ecosystem and consider the complementary of the different tools have something creating value and not have a way to consider all the players as competitors and with which we don’t want to create those relationships that are useful and valuable to enrich and create, advanced, understanding of audiences. And sometimes it’s not possible to create product integration between different methodologies. So sometimes you can do that in an easy way, with one click button. Sometimes you just need to build the right methodologies and you have to challenge your vendors if they have the ability to create the methodology to breach one audience from one tool to another. And if they have built that track record and that approved approach that generates this level of confidence in bridging the tools and enriching the understanding of an audience.
And the last point, it’s about actionability, because, sometimes, marketeers in the last mile, which is the data interpretation in translating the insights from the audience intelligence tool into an action plan could be a challenge. So, sometimes it’s because it’s because marketers, for instance, just want to generate insights that confirm their own assumptions. And in that case, if the results are not going in their direction, it could create a challenge to reinforce their story. But sometimes it’s also because, for instance, They have not really understood what are the strengths and weaknesses of the different methodologies and how they have to use those different tools and methodologies in order to be able to extract, what really matters and what is powerful in each of these different solution in order to to build that 200 degrees, understanding of an audience.
And, I think that to overcome that situation, it’s really important to also identify vendors who is involved from the beginning of the relationship with the customer in doing the right onboarding, in investing time in doing training and enabling the customers to share their main use case, their expectation, and even someone – a success manager, which is monitoring and supporting, the customers in their ability to achieve their use case, building value and elevating their expertise in using audience intelligence. And, if this early stage is done in a good way in a lot of cases, in many cases, the customer is fully autonomous and is building its own expertise without requesting additional support.
But it’s fundamental to have a good onboarding program from the beginning and build a success plan to understand what is the objective of using audience intelligence in order to be sure that it’s generating the end value and supporting business outcomes.
Mike Allton: Yeah. And to your point that the interoperability aspect is key. If you’re using, or your clients are using a CRM that doesn’t have any capability to integrate with other platforms, then you’re blind essentially to what the data is in there.
Robin Dimond: Real quickly: Can you just tell us how agencies, cause you brought this up and you said, Hey, they can manipulate the facts, but how can agencies utilize these insights to win new business or enhance new other clients?
Jonathann Mingoia: With our customer agency, we are observing that they are using our different tools, on three different tracks.
- The first and the most important one is—as you have mentioned—when they are faced with competition and generally timing to provide deliverables when there is competition is super short, the agency wants to demonstrate their full capabilities in terms of audience segmentation and audience understanding because now it’s not a nice way. It’s a must-have to understand audiences in the different segments. So this type of solution is fitting. Perfectly for that need due to the fact that it’s cost-efficient, [and] it enables a high level of flexibility to get different angles of analysis of any audience. And in 50 minutes, you can get a full profiling of any kind of audience. And, on the fun side, brand lovers, people engage with the different owned channels, with customers that are already connected with the brand and it enables them to create ID cards, identity cards. Super easily by highlighting what are the main elements that describe any segment. So it’s super powerful for pitching. Also, during the competition, you can say, Oh, you have seen what we can provide as insight based on open audiences because at that moment you do not have access to the first-party data of the customer. But when we create that you can get also this type of output with your first-party audiences, website visitors, people behind your social channels, and people who are visiting your website, and you can, and we will use, the same lens of analysis that the one you are, observing at this moment in this presentation and you could compare for a different type of audiences, apples, and apples. So it’s quite powerful to use the same lens of analysis across different types of audiences.
- The second track is about improving operations. And you have also mentioned that point. And indeed, when an agency is working on creating recommendations for campaign, content, and architecture building, having access to those insights will help to refine the targeting, refine the messaging, improve the media selection, select the right partners, and at that moment they have access to the first-party audiences. So you can build that knowledge from a genuine audience involved and buying the product of the company. And in that case, you elaborate a strategy for cross, and upselling, improving retention, but also reaching lookalike audiences, online or offline from the insights you are gathering from those profiling based on first-party data online. So it’s super powerful.
- The third area of value is we have observed this last month, more and more agencies were, extending in a certain way, their services by adding consulting offerings, by lighting the ability they have to generate deep research, deep research in a cost-efficient way about brand lovers, about consumers, about any kind of target group. So it’s also a way to extend the capabilities they have on top of building strategies is building strategy and providing a deep understanding of those audiences. That will be the foundation of that audience strategy layer.
I think that we are in a very key momentum in the marketing area where the marketing era where indeed the marketing ultra effectiveness is becoming key. The winners will be those who will be able to put audience intelligence, in the world of any operation innovation, positioning to create more impact and results with less effort and investment.
Mike Allton: Love it. So we started with Apple intelligence and we’ve ended with audience intelligence.
Thank you so much. This has been so interesting and informative.
Thank you for listening to another episode of the MarTech Show, hosted by Robin Dimond and Mike Allton, powered by Agorapulse, the number one rated social media management solution, which you can learn more about at agorapulse.com.
If you want to make sure you’re part of our audience during live weekly broadcasts, take a look at our calendar at agorapulse.com/calendar, or click the subscribe button in your email. Once you register for any of these events, is there a particular tool or topic you’d like to see us talk about, or perhaps you think your solution should be featured? Email mike@agorapulse.com.
