Dec 3, 2024

Gen AI Project Automation - exploring the future for agencies and clients

True Gen AI disruption reinvents the whole journey

Gen AI Project Automation - exploring the future for agencies and clients

Hi!

It was great to have the opportunity to join Kristi Zuhlke for one of her ‘Coffee with Kristi’ webinars last week.

For those that don’t know Kristi, she’s the highly successful entrepreneur and founder of Knowledge Hound and general insights industry innovator. She’s connected with a fantastic community of senior client-side insight, brand and innovation folks who I was lucky enough to get a chance to speak to a few days ago.

Some of you from the webinar signed up to this newsletter - welcome.

I had a great time talking about the Gen AI and automating the insight, strategy and innovation agency process and the potential impact on our industry. The 30 minutes seemed to fly by and I perhaps didn’t get to dig into the detail in the time we had.

So I thought I would give an overview of the key highlights here, as well as dig deeper in some places I couldn’t get to in the webinar. Some elements I’ve touched on in previous newsletters, but I’ll pull all the pieces together and present it as a view on the future of our industry, based on the little that we know today (so much is changing!)

Here goes…

True Gen AI disruption reinvents the whole journey (not small silo’d parts of it)

Most (if not all) of the Gen AI innovation you will see around you is incremental. It builds on and evolves existing processes and tools, perhaps making them more productive, but ultimately doing little to unleash the truly transformative potential of Gen AI for how we work.

This is a common theme in emerging technologies, we’ve seen it time and time again. It’s not until we step away from the mindset and assumptions that have shaped the past that we can fully embrace what’s possible in the future, if we work differently.

It doesn’t mean that there’s anything wrong with these incremental applications of AI. I believe we should all be exploring and experimenting with all kinds of solutions.

But adaption of Gen AI needs to be a two speed journey. On the one hand explore tools that boost the productivity of people and their existing processes and tools. I think it’s great for companies to encourage all of their teams to start using tools like ChatGPT and explore where and how they add value.

But this is not enough, we should all be exploring more transformational uses of Gen AI and how it can create a step-change in the way we work.

That’s why at ONE Strategy Studio we’ve chosen to set ourselves a ‘moonshot’ goal…

100% Automated Gen AI insight, strategy and innovation projects that are as good as, or better than projects delivered by traditional agencies

It’s not about removing humans from the process completely. We still believe in a human journey of discovery. We are an agency ourselves, and we provide a service.

But by setting ourselves this goal of 100% automation we are constantly challenging the way things are done and pushing the boundaries of what Gan AI can do.

Gen AI Automation is happening right now (insight, thinking & creativity - not just process)

It’s easy to imagine that Gen AI automation that can replicate complex human journeys of analysis, insight, strategic thinking and creativity is a long way off. Something theoretically possible but not to think about now. Especially if all you’ve seen is some simple ChatGPT prompting that’s given uninspiring results.

When we started ONE Strategy Studio we thought achieving this moonshot goal was going to be a long journey. And it will be a long journey, there’s much we still we want to achieve. But we were amazed by what’s possible after such a short period of time

100% automation is here already… Now.

In the last couple of months we’ve run over 30 fully automated insight, strategy and innovation projects for a dozen clients across a range of categories and live brand challenges. And we’ve been told it’s as good as, and in some ways better than, projects run by human agencies.

Except these projects are done in as little as an hour, with the ability to iterate… to re-brief and refine and optimise results as much as you need.

This isn’t about automating just process, it’s automating the journey of research, analytics, insight, strategic thinking and creativity that until recently has not been possible with machines. Which is hard to imagine or believe if you haven’t seen it.

Which is why we’ve found sharing a case study is the only way of demonstrating to our clients what’s possible - bringing it to life. If you’re interested in hearing more watch this video - and get in touch - jonathan@onestrategystudio.com

If this level of automation is possible now, imagine where we will be in 6 months, or two years or 5 years? Whilst none of us really know what the impact of Gen AI will be it’s probably time to start planning for some significant changes.

Gen AI automated projects aren’t just as good as a human agency project, in some ways they are better

When we first started we set ourselves what felt like a truly challenging goal, create automated projects as good as those from a traditional insight, strategy or innovation agency. It was our clients that suggested in some ways our projects went a step further and are better. For example…

  • We can go WIDER & DEEPER in terms of sources of insight
  • We can be more OBJECTIVE with how we approach each challenge
  • The AI models fuse BEST IN CLASS EXPERTISE in one place
  • CONSUMER, CULTURAL & EXPERT sources are integrated seamlessly
  • There's a relentless CONSISTENCY & THOROUGHNESS
  • We can get to NEW THINKING that traditional approaches don't get to
  • We can ITERATE the results instantly by changing brief or focus

Here’s some more detail on these points if you want to have a dig…

<insert link>

Gen AI LLMs don’t need to get better to make this automation possible (but they will get better)

Some clients we present to are understandably sceptical at first (certainly before they see a case study or live project). Their only experience of Gen AI has been Large Language Models (LLMs) like ChatGPT or similar. Experience that suggests that current models have significant limitations that would make automation of agency projects impossible.

But these chat interfaces hide the potential of what’s possible with the underlying models today. Potential that it’s possible to release when you use an API that allows you to interact directly with the LLM and overlay your own models and frameworks.

I call it ‘ladder logic’… the more steps up the ladder you climb, the easier it is to see what’s possible just with the generation of LLM models we have now. These steps go beyond asking great questions and perfecting prompting techniques to cover areas like sequencing, curated data sourcing, building frameworks, fine tuning results, journey automation, model collaboration, reasoning steps and model autonomy.

Her’s a link to a previous article I’ve written that goes into this in a little more detail…

It’s still a human journey of discovery (and critical thinking is key)

“If all this is possible, what does that mean for the role of us poor humans?”

I often get asked this, just after I’ve shared a couple of case studies of 100% automated projects. People are blown away and can see the potential impact Gen AI automation can have on our industry.

But the truth is, however much we automate the process, insight, strategy and innovation will always be a human journey of discovery. We just have to accept that our understanding of what parts of the journey remains human will evolve.

We’re only at the very start of this evolution so it’s hard to know where the boundaries between machine and human will be. But for now these are the common themes that we see come up again and again in our discussions. Areas of critical thinking that will likely remain at the heart of human expertise that will go hand in hand with Gen AI automation…

  1. Asking deeper questions & challenging assumptions
  2. Having understanding of and empathy for the human realities
  3. Making the truly step change creative breakthroughs
  4. Understanding commercial context & application
  5. Creating and nurturing human relationships
  6. Navigating organisations, aligning teams and making stuff happen

Here’s a little more on some of these…

The agency business model will need to evolve (maybe that’s a good thing)

What all of this suggests to me is that Gen AI automation will create a transition whereby more and more human time will shift from the process of ‘getting stuff done’ (and by ‘stuff’ I mean process, research, analytics, strategic thinking and creativity) to adding value through these various aspects of critical thinking.

Right now many agencies already add real value for their clients through critical thinking. Times when they challenge, refocus the brief, get to game changing insight & creativity and help navigate the human and organisational barriers to change.

But this is often a relatively small part of what agencies spend their time doing for their clients… and it’s typically not where they make their money. They spend most of their time getting project stuff done and that’s typically where they get their return on their critical thinking investment.

When high level tasks such as analysis, strategic evaluation, creativity and storytelling are automated the picture is flipped. Agencies will need to spend most of their time being a critical thinking change partner and have to find a business model that rewards them for doing this.

Agencies… Now and future…

For some agencies this will be a bigger shift than others. Smaller agencies with senior teams selling process-agnostic consulting may see little change. But those agencies that have scaled based on a small senior team selling large, repeatable and costly process delivered by a pyramid agency structure that gets paid for getting stuff done may struggle to transition.

Most agencies don’t make their money from where they add value. Gen AI project automation could put their business model under a lot of pressure.

But is that a bad thing?…

... Don't get me wrong, I've always loved working in insight, brand strategy and innovation agencies - it’s what I've done all my working life and it's a great industry.

But I've never been a big fan of the typical agency business model...
- Pyramid business shapes that put pressure on junior staff
- Profits only possible when overworking the team
- Day rate charging instead of value charging
- Profiting from process when the true value comes from thinking
... etc

I'm not saying that all of these issues are experienced by all agencies - but I would guess that most agencies experience some of these pressures some of the time.

Even a really successful agency can be like a duck - seemingly gliding through the water with serenity, when in reality they're paddling away furiously beneath the surface desperate to stay afloat.

I'd like to think that Gen AI automation can disrupt the agency world in a good way. It's an opportunity not just to rethink how we work or what we deliver to our clients, it’s a chance to rethink and rebuild the business model as well.

For clients these changes are about freedom and empowerment (but it will take time and effort to get there)

Ultimately we see Gen AI automation freeing up time for client side insight and brand teams to be less focussed on getting project stuff done, and more focussed on adding the critical thinking. More time for…

  • Asking deeper questions & challenging assumptions
  • Having understanding of and empathy for the human realities
  • Making the truly step change creative breakthroughs
  • Understanding commercial context & application
  • Creating and nurturing human relationships
  • Navigating organisations, aligning teams and making stuff happen

That’s got to be a motivating idea, right?

And it’s not about removing agency partners from your lives. After all, we’re an agency too, just a different kind of one. But we see who your agency partners are, and where and how they add value transforming over time.

The nature of project journeys well change too. it’s hard to predict exactly. But here’s some first thoughts on that one based on what we’ve learned so far...

  • Projects will move at the speed of the organisation, not delayed by process
  • Journeys that would typically take 8 to 12 weeks or more can be done in days
  • It will be cheaper - with fewer (or no) depth / breadth trade-offs
  • All internal projects can have a robust high value process, not just the big ones
  • You will have deeper understanding, insights and ideas right from project kick off
  • The journey will be far more iterative and flexible
  • Consumer research comes later - more focussed, delivering more value
  • The need for a well considered human journey of discovery is as vital as ever


That’s it for now. It’s just a glimpse of what’s happening and what’s potentially to come. it could be a hugely transformational couple of years and we don’t know what will happen. But I’m an optimist, I like to think Gen AI and project automation will transform our industry for the better.

Have a great week!

We’re going to continue to explore and push the boundaries 100% Virtual insight, strategy and innovation projects can deliver. The possibilities are so exciting.

If you want to know more about ONE Strategy Studio and our vision for 100% automated Gen AI projects in the insight, brand strategy and innovation space then please do get in touch.

Jonathan

From the first time we connected to today your work simply blows me away, it’s such an amazing proposition in terms of quality and time saving… fantastic and honestly feels like the future is here today.

Daniel ChaDwick
CMI Director - UNILEVER