Humanizing Data in the Age of AI

Because of AI & Big Data, the Insights profession, and the skills that insights practitioners bring to the table will become even more critical in organizations than ever before.

We are all currently swimming (drowning) in a world of numbers. The average human produces around 1.7 gigabytes of data every second. Your phone, your smartwatch, your emails, your credit card, all adding to what is already a huge data set. 

Are we at peak data? Not even close. If you look at all the available data in the world today, around 90% was produced in just the last two years.

[related-article title="Big Data, Small Data, or Both?" url="https://www.greenbook.org/insights/research-methodologies/big-data-small-data-or-both"]

So, what does this all mean for Insights Professionals whose job is focused on data? Good news surely, more data means more insight, right? But of course, it isn’t this simple. A recent Dell Technologies study ((Article 1) found 70% of companies are gathering data faster than they can use it. With all this data comes several risks.

1. Over Interpretation: Companies lean into this data ocean and the interpretation takes precedent over the actual voice/experience of the client or customer. We are all data experts now apparently and so this leads to wrong conclusions (e.g., correlation vs. causation) and what follows is poor decision making.

2. Too Hard Basket: For many companies it becomes too hard to use the data and so decision-making falls back to gut feel or just “we will replicate what worked before” and the true voice of the customer is lost along the way.

3. Trust: We are now living in an era of mistrust. Now that we are in the age of AI deepfakes, Fake News, AI hallucinations (where it references non-existent news articles (Article 2)), executives are now wondering what can be trusted.

I recently presented a very compelling chart that highlighted a strong correlation between a clients view of a bank’s digital offerings and their likelihood to increase business with that bank. The first comment from the global head of the department was “yes but is that what clients are really saying”. Interpretation: The data looks good, but can we get confirmation from actual humans.

This is where Insights professionals come in. While AI excels at processing vast amounts of information and identifying patterns, it lacks the essential human touch required to understand the emotions, motivations, and nuances behind the data. Bridging the gap between raw data and meaningful human experiences is a key skill and the value of this capability will continue to increase as AI potentially drives the dehumanization of data.

The ability to persuade senior management (or any stakeholder) with human-driven insights is invaluable in driving insights into action. By crafting compelling narratives that connect to emotions and resonate with stakeholders on a personal level, insights professionals can garner support for initiatives.

Some ideas on how to do this.

1. Make Them Feel: Get internal stake holders to read out loud relevant customer comments that support your insight to help land your message. I ran a session with internal stakeholders called “Mean Tweets” where we printed out relevant, funny, and often very blunt verbatim and asked members of the team to read them out with feeling. We had some Oscar nomination worthy performances as the team tried to one up each other with their passion. A senior leader in the room mentioned that hearing the customers frustration made them feel “quite sad”. It also made them want to fix the problems because now they had empathy for the pain the clients were feeling.

2. Blow It Up: Take sample data and extrapolate it to your total population: You will likely find that 900 customers complaining about a particular issue may not be compelling if you have hundreds of thousands of customers, but if you extrapolate this number, it can suddenly become tens of thousands of customers potentially facing the same issue. Let’s assume the 900 comments represent 7% of your verbatim responses and your total monthly active customer population is 500,000. A quick calculation could suggest that up to 35,000 customers experience the same issue every month. They just don’t all complain. You are more likely to get action with 35,000 unhappy customers vs. just 900.

3. Frustrate Them: Make your stakeholders be the customer: make them actually do the thing that is frustrating, shop on a budget, find out what is really covered in an insurance policy, complete a task on your website. Make them feel the pain a customer feels.

4. Surprise Them: Add a surprise twist to your insights presentations. A surprise or a counter intuitive perspective can be a memorable moment that will stick with your stakeholders. In a presentation to three internal (siloed) teams I shared the shopping journeys of three customers and their pain points. I then introduced the stakeholders to a single customer who proceeded to share that they were the three customers and shopped all the journeys and felt all the pain points. The surprise made the teams think more collaboratively about how they could serve the “one customer” rather than focussing just on their silo.

5. Make Them Remember: Treat your insights presentation like an advertising campaign and make it memorable. Find a tag line that links the insight to the human experience. During a challenging platform upgrade our clients were struggling to transition to the new platform as it had changed significantly. The new platform had many more features and functions, but change is not easy. We were able to show that while clients initially struggled, their NPS scores rebounded and ended up higher than when they were using the old platform. I coined the term “The Pain of Change” which became a commonly shared phrase to help internal stakeholders understand that we should expect frustration, but it would be short lived.

There are many more ways to humanize your data, but the key is to make your stakeholders feel. The word "emotion" is an adaptation of the French word émouvoir, which means "to stir up" or “to move”. Memorable emotional messages crafted from human-driven insights have a unique power to inspire action.

We want to move our stakeholders both in feeling and in action. Make Humanizing  data your superpower and become indispensable in the age of AI.

1. https://www.dell.com/en-us/dt/perspectives/data-paradox.htm

2. https://www.theguardian.com/commentisfree/2023/apr/06/ai-chatgpt-guardian-technology-risks-fake-article