Better Product Management: Goals and not Flagpoles

Getting a sign off and blessing on a project

Requiring blessings and approvals delays hitting goals and devalues customer validated learning

As a manager, do you feel the need to control every resource and every action of your employees?  Do you want to know everything that is going on–and not just whether goals are being hit?   Do you put in place requirements for sign offs? If so, you’re likely get suboptimal performance from your team.

Often when mangers seek to control a process or know what’s going on, they introduce a sign off requirement.  Time spent getting approvals delivers little value unless it results in very meaningful change.  Instead, this act of ‘running things up the flagpole’ and ‘getting the Pope’s blessing’ mostly delivers value to the micro-manager.  Even if going up the flagpole changes the result, it’s often highly likely that both ideas we’re right–and no customer validated learning occurs to improve future decision making.

Recently a new product management process was provided to our team.  If someone wanted to start work on fixing a high priority bug–even one that affected 50% of online revenue, they would need to get 2 sign offs and alert a third person before work could begin.  This “flagpole” upon which a new activity needed to be run up and approved–provided transparency and control over resource allocation.  However, it means that bugs stay on the website longer, more time is spent getting sign offs, and less time is spent delivering value.

Another example is spending hours getting  approvals for website copy.  After 6 people spent about 6 hours each on copy, another 9 people spent two hours and a half hours in a room approving the copy.  Over 50 hours were spent on copy. The time would be better spent ensuring there was an A/B testing platform and regular usability testing.

Even a customer service organization can create the flagpole problem.  For example, often customer service representatives are required to get approvals for customer refunds.

A high performing company –without flagpoles–could solicit ideas in under an hour (instead of 50 hours), a/b test a few variations, get qualitative insights from usability studies and customer service feedback, and adapt on the go.   Ideas for copy from the 9 people could come via informal 5 minute meetings or emails.  The product manager would be learning faster, the results would be better and more repeatable, and 45 hours would be saved.  Another process done at one public high growth technology company is to have a daily 30 minute meeting with the most senior marketing person where they can provide input on anything launching–and it’s clear that they’re the gate preventing launch so it usually happens in a few days.

Perhaps most importantly, a flagpole process is not scalable.  A high performing 300 person or 1,000 person company should have too many website updates launching each week for 9 people to sit in a room debating for hours on end.  While it’s helpful to have some tone and style guidelines that fit the brand, long meetings for copy approval are not sustainable.

How do you avoid flagpoles?  Give the product manager measurable goals that are NOT just launching features.  The primary question is where are things tracking toward the goal, and what are the next few things being tried to hit the goal. With goals, it also becomes clear what matters more.  For example, if the goal is a conversion rate, the manager can focus on the emailed or verbal copy suggestions that will be seen by 95% of the users–where someone else in an approval meeting might spend precious time on the copy only seen by 5% of users.  And the answer is which copy converted better, not which sounded better (assuming it’s on brand).  With a goal of a conversion rate, the manager can also decide if the bug fix is worth the engineering time and disruption to hitting their goal. Of course, goals are not everything–for example, a product that grows revenue in the short term might hurt the brand and revenue over the long term–but they are very helpful.

Even operations teams can try to remove flagpoles.  For example, to remove flagpoles in customer service, consider allowing representatives to issues up to a certain dollar amount (e.g., $100) to streamline operations (credit to Tim Ferriss for this idea).

As you run your company, watch out for new processes that require approvals, as they can meaningfully slow down progress towards goals.

Did you like this? Share it:

Watch out, Founders! Early decisions lead to awful user generated content

Online communities have a tough time getting started. There is a chicken and egg problem. Who wants to join a community that doesn’t exist? To solve this dilemma, Yelp’s CEO openly admits that early on the company decided to pay reviewers and now pays “community managers” in major cities. A second founding decision was to focus on getting the most reviews possible with its user rewards system.

Yelp’s founding community in any city is those who are willing to contribute reviews for low pay (i’m assuming rates did not attract top writers and bloggers…because it didn’t). For example, for many moderate and expensive restaurants, reviews often cite discounted Restaurant Week specials. Yelp is now challenged with broadening its set of active contributors in order to get broadly applicable content for the larger audience it is attracting.

The second original decision founders made was to encourage the community to review often. Online Karma awards are given for Firsts, and users compete to see who can have the most reviews or Firsts. As the community prefers to be heard rather than rated, Yelp accepts reviews about old meals or reviews that focus on the circumstances of the meal. Sadly, this has led to a proliferation of low quality reviews for restaurants, a category where one size does NOT fit all.

I think Yelp has a lot of potential (17 million unique visitors a month is already impressive), but right now the main feature that helps me is the mashup of google maps, yellow pages, and pictures.

Did you like this? Share it: