Looking Backwards vs. Working Backwards
Are you looking to drive more value from your data or implement high-impact AI solutions? Innovating in AI is not merely about having the most advanced algorithms or enormous data sets but understanding how these technologies can solve real-world problems and deliver customer value.
Too often, companies look backward at what was done before or what customers preferred in the past and then copy-paste it into their organization. The challenge is that context matters — what worked in one use case may not work again.
Instead, look to the customer need first in what Amazon calls their ‘Working Backwards’ process. This approach begins with the customer and works backward to the technology. Many books have been written on the details, but here are a few bullets to remember next time you want to innovate with #AI.
1. Start with the Customer
The first step is to imagine the ideal customer experience - stripping away friction and current frustrations. Don’t just think about what is possible with current technology, but envision what would be perfect for the customer, even if it seems impossible with current technology.
2. Write the Press Release
Next, write a press release announcing the final product (yes, before you start coding). This practice forces you to distill your idea into a simple, customer-oriented value proposition. It also forces you to consider how the solution will be communicated and perceived in the market. You’ll need to be crystal clear on WHO your customer is and WHAT problem you’re solving so that the solution will resonate and inspire them to adopt your innovation. Being different isn’t enough — lean on the power of AI and imagine a 10x solution that is seamless, intuitive, and enjoyable for customers.
3. Test your Assumptions
Share your future-forward press release drafts with key stakeholders early and often. Writing the press release demands discipline and thick skin - you’ll need to get comfortable feeling uncomfortable. But, even if your genius idea gets ripped to shreds at this stage, it’s better to front-load the discomfort than to build something no one wants. A secondary benefit is this practice helps avoid any surprises or misunderstandings with teams who will be critical to the build or roll-out of the solution.
4. Start Building
Finally, it’s time to design the technology. With a clear understanding of what customers want and how they’ll interact with the solution, you can start building the AI solution. Combine the Working Backwards approach with Agile/Lean engineering methods to continue enhancing your press release as you get feedback.
So, next time you want to innovate with AI, remember to look forward, work backwards, and create a future that your customers will embrace with open arms. Don’t let the past dictate your path to success with AI. Ground your next big idea in the customer’s needs and ensure that solutions are not just technologically advanced but also customer-centric.
