Securing a budget for AI projects

How can you secure a budget for exploratory AI data projects without a business case or ROI?

By Helen Tanner

LinkedIn, Twitter, all the social media channels, and even the tabloids are talking about how AI is going to change to the world. And how, in the business world, success will come down to whoever grabs the AI toys quickly and creates a competitive advantage, will win. So, you’d think businesses would be throwing money around to invest in AI, wouldn’t you?

Nope. Speaking to Chief Data Officers, CTOs, Heads of Data and other data and technology leaders, budget is as hard to come by as ever. Particularly in these tough economic times where many businesses are behind budget and reducing costs. It’s a tough market and money is tight.

Finding budget for innovation or research projects is even harder because often the ROI, or return on investment, is tricky or even impossible to quantify. Think of the types of projects you might want to explore…

Using ChatGPT or LLMs add value to your business – you can probably define the cost based on number of people you’d want to involve in a project. But the potential revenue increase or cost reduction? Eeek, that’s a tough thing to estimate.
Using AI and machine learning models to forecast something for your customers – again, you can probably calculate the cost based on people’s time. But the potential price that customers might pay for this additional service…? That’s at best, a hypothesis that needs to be tested.

Without a believable ROI or credible business case, it’s really hard to justify spending money on people or tech to innovate. So, how do you secure budget for exploratory AI data projects without a business case or ROI?

OPTION #1 – TIMEBOX A PROJECT & BUILD SOMETHING IN X WEEKS TO GENERATE A BUZZ

With a background in big corporates, I absolutely love a timebox prove-the-concept- project to get new things in under the radar of the bureaucracy police! You might not have loads of money, but you probably do have a team of people…or a person…or yourself. So, come up with a business problem that can be solved with the latest tech and set X weeks as the maximum time to build a prototype.

When I’ve done this, I’ve always combined two skills together on the timebox project:

A commercial product-type person – someone who understands your customers, their pain points, and can identify what they’d value
A tech person – someone who can loves playing with the latest tech and can build things quickly

With this combination you will have the best chance of creating something quick and of value.

Once you’ve built it, show everyone. EVERYONE. Tour it around the business and sell the story that if this only took X weeks and 2 people, imagine what this business could build if we worked together. Create such a buzz so that at least one senior stakeholder says “I want one” and then work with them to find the budget.

OPTION #2 – USE YOUR PROJECT TO PROVE YOU HAVE AN INNOVATION-LED CULTURE

Truly innovative businesses don’t wait for budget approval or stakeholder agreement before experimenting with the latest technology. They experiment all of the time. They encourage people to play. They support people to try new things. They know most ideas won’t work and that most projects will fail, and that’s OK. They have an innovation culture.

So, can you use your project to prove to the business that they should have a culture where innovation is prioritised. Can you champion innovation as an asset for your business. And this project is just the start, to prove the concept and win people over.

OPTION #3 – CREATE THE BEST GUESS ROI OR BUSINESS CASE YOU CAN

If options 2 and 3 sound crazy, and you need to accept your stakeholders need to see an ROI or business case otherwise nothing will ever happen…I feel you. In which case, let’s get real and make our best case guesstimates:

A. COST ESTIMATES – these might include:

People costs eg X people x £X salary for X weeks/months
Software costs eg £X licence costs
Consultancy costs eg £X to bring in a data expert/s to fast-track the internal team
Training costs eg £X for specialist courses
& so on

B. INCOME ESTIMATES – these might include:

Additional revenue eg £X or X% extra on £x per customer for X customers per year
Grants – perhaps you can acquire a grant to cover some/all of the costs
R&D tax credits – perhaps the project can form part of a tax-efficient strategy
Co-investment – perhaps a partner or client will jointly fund innovation projects
& so on

C. INCOME – COST ESTIMATES – A – B = C!

Position it as your best guess ROI, make sure all your numbers are backed up, and go pitch to the budget-holders.

OPTION #4 – USE AI TO WIN YOU PR & MARKETING OPPORTUNITIES

If none of these options are appealing to you, or perhaps you’ve tried all 3 already, then perhaps it’s all about marketing baby?!

Perhaps you need to be totally transparent about the project being all about generating a marketing buzz…

Get some press coverage
Showcase a cool interactive demo at your next event
Position your business as the only to offer AI-powered services
& so on

Sometimes, for some businesses, in some markets, where everything is commoditised…standing out from the crowd in the press is a huge coup.

Want some help coming up with AI use cases?

Want some help building a business case for AI?

Want some help creating a data-led innovation-enabled culture?

Book a call