September 2, 2022
How would you beat AI & Machine Learning using Jobs-To-Be-Done?
Google, Apple, Microsoft, Amazon, and every big company in China has been investing heavily in artificial intelligence (AI) and machine learning (ML). Across the world, significant capital is being invested in these new and improving technologies. They better be worth it; there better be a return on that investment, not just for global giants but also for your company if you are planning on investing in AI and ML.
If you are on a product marketing or sales team, how should you use artificial intelligence and machine learning to accelerate your growth? Should you use ML or AI in product development at all?
Answering these questions can be a challenge with a lot of roadmapping tools. It's challenging because the only reason to leverage AI or ML on your roadmap or to put anything on your roadmap is that it creates value for the customer. Most roadmapping tools don’t allow you to tie the two together; they don’t give you visibility into how the new solution will deliver greater customer value. Jobs-to-be-Done can help you achieve this because it ties your product roadmap to your customer’s job-to-be-done.
Using Jobs-to-be-Done to figure out where AI and ML fit into your roadmap
You’re definitely going to be under pressure to put AI and ML on your product development roadmap. If you aren’t already answering various forms of the question, “How are you using artificial intelligence in product strategy?” you soon will be.
Back when video was just about becoming the next best thing, Facebook had launched a bunch of video features that were improving the number of visits and all their usage metrics. So the question for everybody who worked on a product team, particularly in internet media, was, “What are you doing with video?” The same situation played out with mobile phones: When the iPhone happened, and the app store got really big, it became about, “What are you doing with mobile?” This question comes up to product managers all the time, every time a new piece of technology becomes popular. You better have a ready answer for “How do you plan to leverage machine learning in your product roadmap?”
However, the more pertinent question for product teams is, “How can X new technology - it doesn’t matter if it's AI and ML, or just an algorithm or some other solution - create value for the customer?” That’s the only reason to use ML or AI in product development. Value means helping the customer solve whatever struggle or problem that they have or enabling them to attain a specific goal, what we call their job-to-be-done.
When you use your customer’s JTBD (jobs-to-be-done) as a focus area or as a compass, it becomes easier to figure out whether a new solution is going to bring speed and accuracy to them getting the job done because that’s the definition of creating value. You may not need artificial intelligence in product development if the way the customer is getting the job done today is still incredibly manual. If your competition is manually creating presentations and using spreadsheets, perhaps you do not need to use AI in your product strategy to create value. If that’s the case, its really good news: You don’t need to learn rocket science to accelerate your growth.
Before you dismiss the idea of AI in product development entirely, however, keep this in mind: If someone (not necessarily your present competition) comes along and gets your customer’s job done faster and more accurately, customers will choose their solution. You want to constantly be working towards the fastest and most accurate way of getting the job done for your customer. If using AI in your product strategy helps you find faster and more accurate ways to enable your customer’s goal, then you need to consider it.
Jobs-to-be-Done helps product teams maintain focus amidst exciting new technology
An easy way for you to maintain your focus on working towards speed and accuracy in getting the job done - whether you leverage AI and ML in your product roadmap or not - is to imagine that the solution to your customer’s struggle could happen at the push of a button. Their whole job gets done when they push the button. Their job could be anything - professional or personal. Getting to a destination on time, living with diabetes, setting a mood with music, getting a baby to sleep at night… anything, so long as it is a job that does not mention any product (streaming music is not a job-to-be-done because it mentions a product).
Once you train focus on this push-of-a-button solution, your product roadmap should always be moving towards achieving it. You survey the landscape of available technology and say “How can we use AI and ML to help our customers push a button and get the job done?”. You focus on the impact of AI on product strategy - on improving customer value.
Almost any customer problem, or what we call their job-to-be-done is actually a system where the customer needs information and then needs to make decisions based on that information, assess if their on track to meet their goal, revise their decision if that is not the case and then conclude the job. Jobs can be anything from optimizing cash flow to getting a baby to sleep at night, and the problem with a lot of jobs is that the information available is incorrect. For example, there are a ton of “get your baby to sleep at night” recommendations out there that absolutely will not get your baby to fall asleep. People are overwhelmed by the number of decisions they need to make on a daily basis across their families, career, hobbies, social life and planning for retirement. And then, they have to deal with incorrect information that makes it harder to make properly informed decisions. Your solution can simply leverage machine learning in your product roadmap so that your ML-backed solution enables better information and, therefore, better decision-making by gathering and verifying information.
Jobs Theory says that customer jobs are stable, but the technology available to do those jobs is going to keep changing. The market opportunity intersects with the technology at the point where the new tech is able to help the customer get their job done faster and more accurately because people will always choose the faster and more accurate solution.
Using Jobs-to-be-Done to choose the right solution
- Accuracy levels
You need to ask what level of accuracy you need for the customer to get their job done and whether using AI in product development is going to enable the desired level of accuracy. For example, you might type “mountain” in search and among all the mountain photos that come up, you might find a photo of a model volcano from a science fair. Not accurate. But also not a big deal. Your customer can move past that photo and simply choose another. But if your solution is categorizing business transactions and you mistakenly categorize a personal expense as a business expense or vice versa, you’re creating big problems for your user, and you need to improve accuracy. Maybe letting your user categorize items themselves is better? It's important to think about the stakes of the problem you are trying to solve to understand how accurate the solution needs to be.
- Feedback, improvement and iteration
The impact of artificial intelligence and ML in product strategy is judged by some (measurable) improvement over the status quo. That way, you can take your solution to your customer and show them the difference between them handling tasks themselves versus your solution doing the job for them. Their feedback and inputs also help your AI or ML solution to become more efficient at what it is doing.
- Customer emotions
There’s an emotional angle to customers adopting solutions that is often overlooked, and that is: the customer needs confidence that their problem is being solved accurately. In the expense tracker example, how does the user feel sure that the hundreds of transactions being categorized are being categorized correctly? Customers need this confidence. Otherwise, they will either not use your product or will do the part of the job that they’re not confident your product is accomplishing accurately, and that means it will take them longer to complete the job. Even the best product road mapping tools lack this emotional angle because they do not bring in this customer empathy (although they are useful in many other ways). However, you can measure these emotions by using a scale where the customer is completely anxious at one end and completely confident at the other end. You see products that achieve this in daily life: Apple and Google Maps have removed the anxiety that comes from getting to new places, for example. If you leverage machine learning in your product roadmap, can you get this emotional angle right?
- Own vs outsourced
Product managers looking to leverage AI and ML in product roadmaps also need to ask whether they should develop their own solution or pick up one that is on the market. If you haven’t got the deep pockets and expertise that Amazon, Google and Apple possess, you could use whatever works (to help you to continue to deliver customer value) in the short term while you build your own in the long term.
To sum up:
You may or may not need ML and AI in product development to deliver the best possible value to your customer today. You do, however, need to keep constant tabs on how you might use them to bring speed and accuracy to solving your customer’s problems over time. That means never losing sight of your customer’s goals and the universe of solutions available to them.
Jobs-to-be-Done can help you make decisions that tie in with delivering better customer value. It helps you decide - for example - whether using artificial intelligence in product strategy - is going to eventually drive more customer value. The thrv app helps you build your product roadmap around your customer’s job-to-be-done and to prioritize your roadmap based on parts of the job where customers struggle the most. It does this by an inbuilt survey mechanism that gives you customer effort scores for the various steps that your customer undertakes when getting the job done, so that you know what customer problems you need to prioritize and can find a suitable solution (maybe AI and ML) to solve them. To learn more about jobs-to-be-done and aligning your roadmap with customer goals, contact thrv today.
Posted by Jay Haynes View all Posts by Jay Haynes