A couple weeks ago, our article on writing design-free requirements triggered some great discussion around requirements and design (also known as “reqs and specs”). What happens when you’re dealing with a website? There are many stakeholders, who are clear about their own goals. Who then turns them into requirements?
A Website Requirements Scenario
Consider the following situation. Your company has a website, and through that website you sell hats- it is an eCommerce haberdashery. You have at least three stakeholders that are important in this scenario:
- Customers (or prospective customers) who interact with the website and purchase products from your company.
- A Merchandiser who is responsible for the content (images and text) that is presented on your website, designed to sell products to customers.
- A product manager who is responsible for sales of a line of products, including through your website.
- A host of others, including SEO experts (responsible for getting traffic to the website), user experience teams responsible for how customers interact with the site, and implementation teams responsible for building different elements (such as the CMS or the customer-data master or the product catalog).
The customer has a straightforward requirement: having a great shopping experience while getting great deals on the perfect products. This is an infinitely deep area for discussion, but to keep this article short(er), the focus will be on the internal players at your haberdashery.
The Product Manager
The product manager is responsible for his product line. In this scenario, he makes high-end, retro-styled hats. The two most important products are the bowler hat and the top hat. The product manager is responsible for the profitability of this retro-chic hat line. The bowler hat sells for $175 with a $75 profit margin, and the top hat sells for $250 with a $125 profit margin. The product manager has a strategic requirement. We’ll use a user story to describe his requirement.
- As a product manager, I need to increase the profits for my retro-chic hats product line, which already dominates the market, so that our stock price will rise.
This requirement is too ambiguous and abstract to be actionable. The product manager can’t really give this to anyone else. What he can do is create an Ishikawa diagram that helps him identify the components of product line profitability, so that he can choose one to focus on. Our haberdashed product manager decides (for this example) that the strategic lever he wants to pull is to change the product mix. He feels that growing the total number of hats sold is not a good strategy – his market is saturated.
Instead, his goal is to convince his customers to buy more top hats – at the expense of buying fewer bowler hats. Each customer who switches from the bowler to the top hat will generate an additional $50 in profit. He further needs to make sure that he doesn’t lose sales to customers who really wouldn’t be happy with anything but a bowler hat. Being measurable is also important. To meet his performance objectives, the product manager realizes he needs to shift the mix from 50/50 to 60/40 (60 of every 100 retro-chic hats sold need to be top hats). His requirement can be rewritten.
- As the retro-chic hats product manager, I need to increase the percentage of top hats sold to 60%, without reducing the total number of hat sales, so that I can increase the profitability of my product line.
OK, that’s measurable. Whatever someone does, you’ll know if it succeeded. But is it actionable? No – it is still ambiguous.
The merchandiser is responsible for finding the best photos, writing the best ad-copy, and determining the marketing messages that are displayed to customers on the website. The merchandiser is held accountable for conversion – the percentage of visitors who come to the site, view the content, and ultimately decide to purchase. Since your company is not dis-functional, the merchandiser has aligned her goals with the product manager.
- As the retro-chic hats merchandiser, I need to convince online customers who would have bought the bowler hat to buy the top hat instead, so that the percentage of top-hats sold is at least 60% of all retro-chic hats.
There’s an implicit choice here – a specification. Here’s another user story that would achieve the same goal.
- As the retro-chic hats merchandiser, I need to increase the conversion rate for customers that view the top hat page from 2% to 3%, so that the percentage of top-hats sold is at least 60% of all retro-chic hats.
Since they are both actionable, but both very different, that works as a good litmus test that the product manager’s requirement, while measurable, was still ambiguous.
There could also be an SEO expert, for whom the story would be “…increase traffic to the top-hats page without decreasing conversion…” A member of the the site design (or user experience) team, could try and improve conversion by creating a more efficient flow (like “one click ordering”) that increases conversion rates – but only enable it for the top hat product.
There are just too many cooks in the kitchen. Every one wants to feed the hungry people, but each specializes in a different dish.
You have at least four viable ways to approach solving the product manager’s goal. You need to pick one before you get the development team involved.
A Matter of Perspective
The merchandiser, SEO expert, and site design teams all have the same goal – support the product manager and achieve a 60/40 mix in retro-chic hat sales. Each of them can develop a reasonable requirement, within their own domain.
To a man with a hammer, every job looks like a nail. The merchandiser is not going to focus on purchase-path redesigns, nor is the SEO expert going to suggest changing from a boring photo of a manikin to a photo of an attractive man laughing with friends over an expensive meal.
The product manager, however, should not be telling the cooks how to make the stew. The product manager (of the product being sold) is not responsible for, and should not be driving the decisions about how to achieve his 60/40 mix goal. The product manager could think of the people in each area as vendors, “selling” alternate solutions to his problem. Each solution has a cost, so the product manager probably can’t just “do them all.” Unfortunately, the product manager is not qualified to pick the right vendor – only to evaluate the promised benefits (from each) at the projected costs (from each) and hope he chose wisely.
A Website Product Manager
The critical need to product managing your website is more visible than ever. Without it, you won’t make good decisions about how to address your real business goals. You need someone who understands the risks, trade-offs, and possibilities inherent in each of the four solution approaches. If your website is large, you may have a portfolio manager for your website, with individual product managers owning areas of the site.
Recognize that not only are your (company’s) customers your customers – but so are your other stakeholders. As a website, you create commerce, you don’t just sell products. You have to help (external) customers have great shopping experiences, while helping (internal) customers meet their goals.
The website product manager works with the retro-chic product manager to understand his goal of achieving a 60/40 mix in sales. Taking into account the big picture from a website perspective (who are our customers, who is the competition, what is our vision), the website product manager has to make a “design decision” and choose one of the solution approaches.
The website product manager is constraining the solution space by selecting one of the hammers. Is this bad? The retro-chic product manager constrained the solution space too (when deciding that a shift in mix was the “right” answer).
Then the how would we solve it, if constrained to solving it this way, user stories have to be developed and given to the implementation team.
You need to have someone acting as a product manager for your website. That person is in the ideal position to constrain the solution approaches to addressing any given stakeholder. Those constraints do limit your options – so you better have a good product manager. It really is no different, however than saying you better have a good general manager, so that the strategy you are supporting is a good one.
What are your interactions like with the folks that own your website? Or are you already a website product manager?