Eric Rems
  • Home
  • Blog

Thoughts on Product and Life

Managing Outcomes vs. Outputs Through Product OKR's

11/2/2020

 
Picture
When building a technology product - whether you're creating an application from scratch, or evolving an existing platform -  it's paramount to have clear goals set, and ways to measure those goals throughout the entire product journey.  

In my experience, without those well-defined goals it’s far too easy to fall into the trap of “shipping product” as a measure of success. 

“We launched 10 new features this quarter. High five!”

Shipping product is an output, not an outcome. If you focus too much on outputs, such as how many features your team shipped in a given quarter, the customer's needs can start to take a backseat. I've seen this scenario play out across several organizations, perhaps you have too...

The Product Manager wants to make sure that their scrum/development team stays busy or else they fear that they'll lose those precious resources to other teams. So, they keep adding and adding to the product backlog and continue launching new features without an outcome in mind to measure success. The result, "feature bloat"...think the clutter of the (old and probably current) Yahoo homepage. On the flip side, think about the clean nature of Google's homepage. That's because Google has always focused on the outcome of "delivering the best search results" to their customers.

Outcomes directly align with what a customer wants and allow the product leaders (Product Managers, Entrepreneurs, etc.) to operate in a more customer-centric manner (this is how the great companies run, think Amazon). Outcomes are what also motivate a team to really empathize with the customer and take ownership of their problem <- this is where the magic happens.

Remember, “Customers don’t want a quarter-inch drill. They want a quarter-inch hole“ - Theodore Levitt

Quarter-inch drill = Output
Quarter-inch hole = Outcome

Now before we get into how OKR's can help us build better products through focusing on outcomes, let's level set by looking at the two things I believe technology products need to do really well: 
​
  1. Solve problems 
  2. Align with organizational goals
Picture
This simple test stands true across the spectrum - from non-profits to multi-national corporations. Even non-profits that launch technology tools need to ensure that they are in tune with the larger organizational goals. If not, you'll have a massive disconnect and precious dollars will be sent down the drain. 

[OKR’s have entered the chat]
OKR’s are a great goal setting framework because they marry a qualitative vision with measurable ways to determine success. They help us product builders manage outcomes and ultimately build world-class products. 
​
On a personal level, the OKR framework has been extremely helpful to me over the last three+ years running a technology startup where we were one of the first entrants in the space. Because the technology itself was so new, conversational AI, the end users (customers) didn't always know what it is they wanted from it + we didn't know what would stick and truly solve a problem for them.  So, we ran a ton of experiments - but, we always made sure we had an outcome in mind when doing so. By having clear outcomes set, it made it easy to cut those products/features that weren't meeting them - and equally important, double down on those that were. 

OKR's also helped make sure that we were aligning with the larger company vision (showcased in the diagram below).
Picture

So what are OKR’s
O = Objectives — “what” it is you are trying to achieve. Simply, the goal.
KR = Key Results — “how” the objective (goal) can be achieved. Also a measuring stick.

They follow this formula:

OKR = 1 Objective + 3x Key Results (note: the # of key results varies by preference, I like to go for 3)

How often do you define OKR’s
For product-focused OKR's I stick to quarterly — this keeps them more tactical and focused. From my personal experience, products evolve rapidly and often. Company-wide OKR’s are often better suited for annual, but it’s ok to also include some annual OKR’s that are product focused. 
​
Writing great objectives
When writing an objective, make sure that it meets the following requirements:

• It’s bold and aspirational
• It’s specific
• It's in alignment with organizational goals

• It's measurable

Writing great key results
Key results serve as the steps needed in order to successfully reach an OKR objective. Effectively, if you meet your key results you’ll have then met your overall objective. 

An analogy that I think paints the picture nicely of what key results are is that they serve as a 'GPS' that helps you navigate towards your goal (objective). Key results should follow the SMART method (Specific, Measurable, Actionable, Relevant, Time-bound).

Let's write an OKR together
But first, let's start by what NOT to do. Here's an example of a poorly written product OKR (then we’ll optimize it)...

O: Improve traffic to our website in Q2 2019
KR 1: Have team write more articles
KR 2: Add new ways for chatbot users to access articles
KR 3: Analyze traffic to website and think of ways to optimize

Off the bat it's obvious that the objective (goal) itself is vague...
What are we comparing the improvement to?
Are we talking about unique users? Total page views? 


The key results are equally as vague. These are not measurable or actionable.

Now, let's improve it...
​
O: Drive 500k unique visitors to themindfultechlifestyle.com in Q2 2019
KR 1: Publish 2 articles per week during Q2
KR 2: Increase call to action buttons within all live chatbots by 25% in Q2
KR 3: Reduce website load-time by 50% as measured with GTMetrix

The objective itself is now clear and measurable. Our goal was to reach 500,000 unique users during Q2 2019 (April - June).  The key results here are equally as detailed and actionable as you can see. These can now individually serve as our GPS as we navigate towards our goal. 

5 Real world OKR's from my company
Below, I've shared a bunch of actual OKR's from my last company to help you frame out your own, enjoy...


O: Increase profitability by 15% in Q3 2019
KR 1: Optimize display advertising through Google Adsense to only show ads with a cost per click greater than $.50
KR 2: Decrease 3rd party platform expenses by 25% 
KR 3: Increase daily article broadcasts by 20%

O: Decrease the number of default responses sent to chatbot users by 25%
KR 1:
 Use Yandex reporting tools to create a weekly report of utterances not understood by all chatbots and have them sent via email to team
KR 2: Divide report equally across team to conduct a weekly review of report
KR 3: Formulate and launch responses for at least 75% of utterances not understood 

​
O: Increase button click through rate (CTR) by 25% across all chatbots by the end of Q4 2019
KR 1: Change all text in call to action buttons to use active language
KR 2: A/B test call to action buttons for each daily broadcast to identify the best results and implement
KR 3: Hire an expert copywriter to refine and optimize all buttons by November 1

O: Launch versions of 3 chatbots on voice channels in Q1 2019
KR 1: Hire a developer with experience on Alexa by Jan 15
KR 2: Create a simple version of MeditateBot, MotivateBot, and StoicBot on Amazon Alexa
KR 3: Cross promote to new channel via call to action buttons in persistent menu of related chatbots on Messenger

O: Get 50 positive testimonials and showcase them on company website in Q4 2018
KR 1: Setup designated email address to receive and review testimonials from users
KR 2: Encourage chatbot user base to send short reviews of their experience to success@themindfultechlab.com with call to action buttons in daily broadcasts in Q4
KR 3: Create a page called “Impact” on company and include the best 50 testimonials

Eric Rems
ericrems.com

    Author

    Sharing my perspective on product building and a little bit on life. 

    Archives

    November 2020
    February 2017
    October 2016
    June 2016
    February 2016
    November 2015

    Categories

    All

    RSS Feed

Eric Rems © 2021
  • Home
  • Blog