I think it’s safe to say that 2024 was an extremely dead year for creative dataviz freelancers and small agencies.

• We see less experimental data visualizations, data viz showpieces or key visuals, and less presence in exhibitions.
• Also on the commercial side, there is a noticeable decline in investments in bespoke data visualization from freelancers and independent studios.
• Formats have changed: the few viral, well-known data visualizations tend to be static images or movies or small, personalized apps rather than elaborate interactive web pieces.
• There is a growing homogeneity in visual styles, with many projects looking similar due to the prevalence of established formats, templated solutions and software defaults.

Why?

What could be reasons? Well, as everywhere — it’s a polycrisis.

• In an economic crisis, everybody plays it safe: This leads to much less enthusiasm for investing in experimental, innovative work or hard to quantify outcomes.
• Off-the-shelf platforms are getting better and better, and are often “good enough”.
• AI is eating our lunch in terms of being the “cool kid” everybody wants to hang out with, but also replacing some need for custom tools.
  • Regardless of if our work can actually be replaced by AI, just the belief it might be possible in one or two years already hinders us.
• Public funding for arts and culture is reduced everywhere.
• The media landscape has changed dramatically. Our frantic social media and generally just increasingly fragmented world favor at-a-glance, shareable, bite-sized content.
• Fact-based communication has partly been replaced by vibes and lore.

• Design landscape is changing: Agencies and freelancers are increasingly brought inhouse, where they actually fulfill crucial business functions. But even there, they often lack the proper influence to affect fundamental change (see also "Product design is lost").

• Industrialized culture in general tends to rely on remixes of known patterns. Just look at Hollywood. We might be part of an industrialized form of culture now, after the gentrification phase in the 2010s.

• Lack of client education: Many clients still undervalue data visualization, seeing it as a commodity rather than a strategic craft. Smaller practitioners spend extra effort convincing clients of the ROI for thoughtful, custom work—efforts bigger firms can skip thanks to established trust. Clients might also demand fast turnarounds or trendy features (e.g., “make it interactive!”) without understanding the complexity, forcing independents into a reactive, rather than proactive, creative mode.

As it seems, we have not explained our workflows, and value add well enough?
I raised similar points already 8 years ago — how much has changed, really?

What works?

So, just to be clear — while we can all think about how we can improve, I think a lot of these factors are beyond our controls and really not dataviz practitioners’ faults at all.

Yet, we have to wonder — what can be done? And how do we adapt? Which approaches do actually work for our clients and audiences?

Let’s look at dataviz for communication first:

I think, there’s an interesting dichotomy at play:

• After COVID, people are now way more used to dashboards and charts. Good!
• But, there’s a even higher expectation for at-a-glance, shareable, bite-sized content that clicks immediately.

I asked around on bluesky for dataviz pieces that went “popular” — pieces that went viral, found wide adoption, opened new audiences, were shared a lot — and compiled some of my favorite responses here: "Popular dataviz".

First of all — I got lots of great, and actually quite varied responses, proving that dataviz in fact be quite successful for getting people’s attention.

But if you squint your eyes a bit, you can see how these popular pieces seem to fall into two classes:

1. Simple, catchy graphics or short animations:

   • Warming stripes
   • Hammer and the Dance / "Flatten the curve"
   • First Image of a Black Hole
   • …
2. Easy-to-personalize “apps”
   • Should I rent or buy
   • Chronotrains
   • Apple Health “Rings”
   • …

So, we really need to have a finger on the pulse for what actually works — and not just gravitate towards complex, intricate looooong scrollytelling pieces, just because we like to build them.

I came to a similar conclusion when I did my Waves of Interest redesign in 2024, based on the 2020 site: I ended up with a much leaner, much more focused site, on single screens, and centered on the key visuals — but also more shallow, less communicative and less nuanced.

On the applied, task-oriented side of data tools:

We need to acknowledge that there are many more ways to create good-enough dynamic charts and charts with little or no coding effort — think PowerBI, Tableau, Google Looker, Grafana, Superset etc. with the additional benefit of integrating directly into existing information ecosystems. So if a good enough chart or report can do the job just fine, there is no need for hiring a specialist who will come up with a custom solution, which will be hard to adjust or maintain long term with inhouse staff.

AI in the form of chat (and notification) interfaces will also replace some “I just need the numbers" tasks”.

But, you know what — that’s fine!
Do we really want to build another standard solution by hand?

There’s still plenty of consulting need for a really good overall data strategy and putting the chosen platform to optimal use, or data visualization guidelines and libraries in this space — but then we talk about strategy and systems design rather than crafting.

Second, as I mention in the post already, if you work as an external consultant on applied tools, you really need to engage deeply with people’s objectives, work processes, visual literacy, data situation etc. Ideally, you work together so closely that your clients co-create parts of the solution. Only then can you make sure your solution actually fits their needs and is also recognized as such!

In order to succeeded in these spaces, you need to clearly state your goals and have a way to check if you reach them.

In the Peak Spotting project, we made a conscious decision to define HEART metrics for every new feature and test if they were adopted as expected. I didn’t use this method anymore in follow-up projects, but now I think we should revisit it! How else can we come up with hard, tangible evaluation of our outcomes?*

So, if you take these things — data strategy, integration with platforms, user-centeredness, scaling beyond your immediate output, metrics — really serious, there’s still a lot of wonderful work to be done and success to be had in this space, both for externals and inhouse staff.

How we’ll bounce back

Creative data visualization is far from dead.

As explained e.g. in this talk, I think we as a community bring a very specific mindset and skillset to the table that is unique and incredibly valuable:

1. Double down on data with an attitude, craft with purpose

   I know data visualization practitioners as typically highly mission-driven, curious and content oriented. We care deeply about what the data has to say, what its limitations are, and how we can make an impact. We bring a journalistic drive scrutiny to any project and, if the stars align, can create that surprising new perspective on the data that suddenly makes everything effortless to see and experience.

   Second, we are the ones who can connect abstract data to human experience, get things on a level where people can actually connect to the information at hand.

   Let’s make sure people learn to appreciate that quality and hire us not only to make things “look fancy”, but understand the value of this approach, involve us already in conceptual and strategy phases — and pay the corresponding fees.

2. Real artists ship

   One thing my clients love about our work, is that we produce tangible outcomes early on, and are just very capable of getting our hands dirty and start building, rather than debating and sketching and layouting and spec’ing and getting approvals before anything actually works. This can-do, let’s-try-it-out attitude, compared with the actual technical and design skills to get a first prototype out and keep iterating on it, is something people usually find very refreshing.

   Dataviz teams are just great for early proof of concepts and prototypes, especially when they rely on drawing insights from actual data.

   The practical downside here: Prototypes and exploratory concepts are typically not paid super well, rather seen as “entrance test” for the big project, with “ACTUAL WORK” 😒 i.e. lots of coding.

3. Make our work last

   We love crafting things ourselves.

   But in order to make the investments worthwhile, rather than treating data visualization as a single-use artifact, we should always also think about how our work can continue to grow after we move on:

   • Choose technologies that are open source and have a good shelf life.
   • Build tools and frameworks that allow others to continue the work.
   • Create modular, adaptable solutions that provide long-term value to clients and organizations.
   • Involve users and other creators early on, so they see how things are made, grow to co-own the results and ideally carry on the torch!

4. Expand the Medium

   Most simple promotional graphics or standard dashboards are not designed by dedicated data viz experts. And maybe that’s totally fine!

   Let’s look ahead — how can we integrate data visualization into broader storytelling formats—games, AR/VR, generative experiences? How can we provide beautiful and helpful ambient information displays? Are the ubiquituous chat formats really our enemies? Or rather an opportunity to rethink dialog-based data experiences? etc etc etc.

The work starts now — let’s open that next chapter 🙂

See also

A crisis of meaning in UX design

The big design freak-out: A generation of design leaders grapple with their future

Product design is lost