In my biased opinion, Tension has been delivering outstanding product and experience strategy since March 2020 — and all we’ve ever shown of our projects publicly are single images, and client logos. That ends today.
We’ve been too busy over the years to truly craft the stories of the products we’ve guided our clients to create. Even though that’s still the case, enough was enough. Two years ago, we crafted the layouts, built the CMS and web templates, only to sit there untouched ever since. Which is both a testament to how busy we’ve been, and also to just how much we prioritize client work over our own.
We thought we’d try a little experiment: we built ourselves an internal ai-powered case-study generator pipeline. Nothing serious, it only took us a few hours to get it functional, and a few more to get it to output draft content that was good enough to work from, in the right tension voice and tone, and in the correct formatting.
Our pipeline takes project documentation like kick off decks, persona documents, journeys, and final strategy docs, and creates interview guides for the specified projects, then interviews us via voice about the projects, branching for more clarification where needed, focussing on what mattered, where the aha! moments were, and why our approach made each project impactful. It then outputs appropriate copy according to our template’s CMS fields for web, as well as long form case studies simultaneously for each project.
Then we pick the images ourselves, add them to the CMS, edit, clean up and add context to the copy and import that directly to the CMS as a CSV, and there you go — we now have our first 10 case studies done and live on our website for the first time in our existence.
Something that we couldn’t get around to for years, done in a matter of days. The next batch of 8 will probably take a day (80% of that time will be for manual project image selection and prep), as we continue refine how we use the workflow.
At one point before this summer, we were supposed to do this manually — human to human interviews for each project to jog the memory, crafting case study content by hand, massaging, editing and creating the various content pieces required.
Mogul
Mogul by MoviePass is the first Hollywood Fantasy league built entirely on the blockchain. Build your studio. Acquire actors, directors & films. Become Mogul.
MoviePass
When MoviePass wanted to build their loyalty platform on-chain, they turned to us to help define an experience that would work for crypto enthusiasts and movie fans alike.
Doppl
We helped Interaptix redesign the experience around their spatial environment explorer app Doppl, which lets you crop elements from the splatted environments at bring them back into your real world spaces.
RMC
What started as a simple truck tracking app quickly became an 100% digital estimating, assignment, ordering, tracking and Customer Service ecosystem for a well-respected concrete company.
Black Talon
In 2024, we worked side-by-side with Black Talon to help imagine a transparent window into each of their customers’ individual attack surface and total risk, including vulnerabilities, threats, and asset and human risks.
Gluxkind App
In 2024, as part of our 4yr celebration, we helped Glüxkind imagine the art of the possible form their companion app, considering the user experience, hardware and software integration and enhancements.
Performatrin
When Performatrin was introducing new product packaging and positioning, we were asked to design a website that reflected their ethos, dedication to natural nutrition, scientific diets, and performance.
Stratus
We were asked to design a new system that would facilitate sharing, offer collaborative workflows, allow for easy archiving and retrieval of individual LiDAR scans in an intuitive way that anyone could use
Eon
Dark tunnels, 2 story trucks rolling by, the risk of falling rocks — all while geologists and geotechs are trying to take detailed scans and measurements of rockfaces. No pressure to get the experience right at all.
UOPX
In 2020, we worked side-by-side with Univeristy of Pheonix to reimagine education in a reality of short attention spans, competing priorities, and a need to self-improve. The goal was simple: make it eas to make yourself better,
View case study
Case study creation is time consuming and can be painstaking work. Again, the image selection and prep is a huge part of the effort as we have to ensure no proprietary client information is presented to the public.
But, it’s not just that the whole process would have taken a very long time (though, it would have), it’s that the initiative never really took off. It sputtered and died before it really even began — as the team (rightfully) focussed on our clients’ projects and needs — leaving our own squarely and perpetually on the back burner. In this case, as in many others, AI not only sped up and simplified the process, it made it more human, more natural in many ways. Yeah, a computer did a lot of the heavy lifting, and executed the interviews, but, the interviews themselves were in fact the most human way of getting the information about these projects, locked away in our minds for sometimes years, ensuring the project details came out as stories, rather than we did x and y and z.
What’s really interesting to me in this scenario, is not that the AI enabled us to produce and release case studies, or that it it reduced the time to produce them, or even that less humans needed to be involved, though all of the above are true, it’s the fact that by considering and understanding how best to use AI in this particular scenario, we were able to do away with a lot of the preconceptions of how a human uses a computer to achieve a goal. It wasn’t so much a this, then that, then this procedural approach, it was a “what’s the outcome I want to achieve, and how can I leverage AI to do it in a way that feels more natural and engaging and still get the result I’m looking for?”.
This is what surprises me about how AI gets leveraged today: most who try to experiment with it are trying to automate exactly what they do today, just so it’s a bit faster, which is an understandable, but complete waste in how to exploit AI. AI isn’t an incremental improvement, like faster networks or a larger hard drive, that simply allows you to do the same thing but faster. AI, at it’s core, allows you to rethink “why” you do things the way you do, challenge the processes, throw away things you’ve become accustomed to doing because that’s what the tech you had available was capable of, or the limited resources, time or money you have. AI allows you to blow up everything you think is standard, or codified, or “it’s just the way we do it” and find new ways of producing by starting with the result, and working backwards. Over the last few months, I’ve been not so secretly questioning everything we do. Not the outcomes or results, but how we get there. “Is this the only way we can do this?”, “Do we really need to do all these things to get these insights”, “What’s the human value, to us or our clients, of this activity?”. I’m not trying to simply speed up, I am trying to find ways that we can spend more time being strategic and creative, while reducing the time requirements on our clients, and deliver better, more interesting results. It’s a difficult process, as it requires you to question the value of what you do, when you have placed so much value on what you do.
The realization I’ve personally had is: it doesn’t matter how you get there, as long as the quality, validity, and craftsmanship is better than it was before. Don’t love the process, love the product.
We continue to leverage AI in new and exciting ways internally to optimize time consuming internal processes that allow us to work faster, smarter and better across the board for our clients — same level of innovation, same level of care and craftsmanship in a fraction of the time it used to take.
Just one example, but I wanted to share it because it’s the perfect spotlight on how tension is continuing to reimagine our own processes and ways of thinking and working, to better serve our clients, in shorter timeframes.
We’ve got a few really wild ideas we’re about to to begin testing (not on paying clients of course) that we think may have the potential to implode one our most time-and human-heavy project-based efforts. The idea itself feels insane and foolish — so we’re probably onto something!
More on that in the future… maybe
Starting something? Maybe you’re building a new product or service from the ground up, or revamping an existing one? We can help.
