Live Blog: Connecting with Customers, Pollyanna Pixton, Agile Development/Better Software Conference East 2013
What do customers want? Well, they want everything. Customers try to give the solutions, but you need to understand their problems. So you have to figure out what they need. Pollyanna Pixton talked about how you can go about finding what will truly bring your customers value.
Learn from your customers consistently and constantly by asking these important questions:
- Who do we serve?
- What do they want and need most?
- What do we provide to help them?
- What is the best way to provide this?
Pollyanna recommends "the billboard test": Ask your customers to explain why people should buy their product by using language that could go on a billboard. It must be concise, quick, and eye-catching!
"Decision filters" can help you focus on what the problems are. When making a decision, answer based on whether it will help your customers achieve their goals and solve their problems. Cascade these decision filters throughout the organization. If you regularly get questions that don't pass the decision filters, find out why they keep getting brought up. Also, integrity matters: Don't just "sort of" pass the decision filters. That won't help you.
Sometimes, we get so focused on our particular responsibilities that we forget we're part of the whole process of building the product, including customer satisfaction at the end. Getting customer feedback as soon as possible is immensely helpful. Give customers a minimal viable product they can use and see if it passes their decision filters and is what they need. Pollyanna says too often, if your product is a layer cake, in the testing phase we give the customers the cake, then the frosting, then the filling. What we should be doing is giving them a cupcake: a smaller but completely functional product.
A month into a project is an ideal timeframe for getting the first feedback. Process that feedback by asking some questions. What have we learned? Has our understanding of the problem changed? Do we need to make adjustments to our decision filters? What can be done later?
Remember, though, that you have to manage your customers' expectations. Don't overcommit to requests if you can't keep those promises. When you tell customers you will do something, do it.
You can make this easier by staying connected with your customers. Create a customer forum. Include all types of customers, make sure all needs are represented, and ask them to help prioritize your backlog. They'll tell you which features are most important!
Keep up by asking for continuous validation. Demo with customers at least once a month, and have customers on site with developers often.
Of course, you will get some pushing back against these proposals. Some typical barriers managers will bring up are that there are too many different types of customers, you don't know who your customers are, and you can't find anyone to respond. It may take some work to find your customers, but they're always out there. Create a focus group, settle on a group of frequent and typical users, or even use your friends and coworkers.
Lots of times, the business contact will tell you, "I talk to the customers, so I already know what they want." No! Information gets lost and filtered as that feedback goes up the chain from customer service to marketing to project manager to program manager. Get feedback straight from the source.
Pollyanna says the best question you can ask to determine customer satisfaction is "How likely are you to recommend this product?" There's controversy over this promoter score method, but it's definitely a quick way to take the pulse of your product.
Make sure your metrics are useful to your customers. Beware of vanity metrics that don't measure anything of value. Good metrics measure progress toward solving the problem. Don't have too many metrics, either—keep it simple with about four or five.
You are responsible and accountable for knowing the value to your customers and to the business. If you don't know the why and the what, you won't know what you're building, and the product won't be any good.