Category Archives: Internet

Shiny new objects rust over time

Many of us have worked in companies that prioritize work on an annual basis. In these organizations, it’s common to invest in a few things each year at the expense of everything else. “This is the year of search,” you might say, because “last year was the year of mobile.” On the surface, this makes total sense – you can’t make big investments in the same capabilities each year. Eventually, everything will have its day in the sun.

In reality, this is a terrible way of prioritizing work, completely ineffective, because it means that – until each program has its day in the sun – it’s essentially left to wither and die. Or to almost die, which is even worse. Don’t you see it? Look there in the corner – there’s site search, begging to be put out of its misery. It had its last big investment two years ago. It knows it’ll be the golden child again next year, but in the meantime, customers are mocking its uselessness, laughing at its broken experience. Poor thing. 

A few years ago, my company addressed this by implementing an Agile, product-focused approach. Within our organization, each product area receives some amount of ongoing funding, which allows the team to make consistent progress throughout the year. The amount of funding may change according to business needs, but the idea of making consistent progress across the business doesn’t.

Take the Search product for example. Several years ago, Search had its day in the sun. Replacing our old system was one of the company’s top priorities, and the project got all the people and money it needed to be successful. Once our new Search was implemented, the big project was over, but the need to make continual progress – to optimize the experience – was still there. (It will always be there.) So although the Search team got smaller and some resources were moved to new strategic priorities, the capability continued to exist – we continued to make it better.

Search is, of course, just one example of a large company initiative – a shiny object – that needs ongoing care and feeding. (I can’t think of a product that doesn’t.) Organizations built on a product management model get this. Before they build new functionality, they’re thinking about how to support it and what the long-term roadmap looks like. Organizations that are entirely project focused, on the other hand, are likely to face poor support and unexpected costs. That’s because they haven’t planned for the inevitable, and because rust never sleeps.

Related: How much of a problem can you buy away?

How much of a problem can you buy away?

My eCommerce team is world class – the best group I’ve worked with, capable of leaping tall buildings in a single bound. From top-to-bottom, we’re defining compelling visions, creating aggressive roadmaps, delivering innovation and quality, testing into winners, and measuring results. Even still, we can’t be experts at everything, and resources are limited.

This means there are times when we need to explore the vast vendor landscape to find tools that can help us solve our problems. We tend to focus in three areas:

  • Things that have been commoditized, because we don’t think they’ll help us differentiate.
  • Things that are really, really hard, because we don’t have the desire (or time) to build or maintain world-class-systems that we can easily buy.
  • Things that are outside of our core competencies, because sometimes the best way to get started with a new technology is to have someone else build it and then take it over.

In each of these areas, there’s a sense we’re buying a problem away, that by leveraging a vendor for ratings and reviews, or personalization, or search, or recommendations, or whatever, we can free our resources to focus on other things. And this is mostly true. But it’s never completely true.

It’s never completely true because there’s no problem that can be bought away completely. Even if we’ve “completely” outsourced a program (like ratings and reviews, for example), we need to define our strategy, manage our vendor, provide customer support and issue escalation, measure the program’s effectiveness, tweak the experience, and defend our budget. Despite our vendor’s promises, these things cannot be outsourced. Not effectively, anyway.

And who would want to hand all of these things over anyway? Vision? Analytics? Roadmap? I may want to leverage a vendor at times, but I never want to hand them my company’s strategy and ask them to drive it.

What does this mean? It means that when we consider buying a problem away, we need to understand that we can very rarely buy away the whole thing. We need to allocate both resources and dollars for the activities listed above, for ongoing support, and for the cost of potentially moving a capability in-house over time. We need to understand that there’s no such things as “hands-off” management or programs that run themselves.

Related: Shiny new objects rust over time