In my last post I spent some time talking about why we care about measuring retention rates, and tried to make the case that retention rate works as a meaningful measure of quality.
In this post I want to look at how a few key metrics for a product, business or service stack up when you combine them. This is an exercise for people who haven’t spent time thinking about these numbers before.
If you’re used to thinking about product metrics, this won’t be new to you.
I built a simple tool to support this exercise. It’s not perfect, but in the spirit of ‘perfect is the enemy of good‘ I’ll share it in it’s current state.
Optimizing for growth isn’t just ‘pouring’ bigger numbers into the top of the ‘funnel‘. You need to get the right mix of results across all of these variables. And if your results for any of these measurable things are too low, your product will have a ‘ceiling’ for how many active users you can have at a single time.
However, if you succeed in optimizing your product or service against all four of these points you can find the kind of growth curve that the start-up world chases after every day. The referrals part in particular is important if you want to turn the ‘funnel’ into a ‘loop’.
Depending on your situation, improving each of these things has varying degrees of difficulty. But importantly they can all be measured, and as you make changes to the thing you are building you can see how your changes impact on each of these metrics. These are things you can optimize for.
But while you can optimize for these things, that doesn’t make it easy.
It still comes down to building things of real value and quality, and helping the right people find those things. And while there are tactics to tweak performance rates against each of these goals, the tactics alone won’t matter without the product being good too.
As an example, Dropbox increased their referral rate by rewarding users with extra storage space for referring their friends. But that tactic only works if people like Dropbox enough to (a) want extra storage space and (b) feel happy recommending the product to their friends.
- Build things of quality
- Optimize them against these measurable goals
At the end of last year, Cassie raised the question of ‘how to measure quality?’ on our metrics mailing list, which is an excellent question. And like the best questions, I come back to it often. So, I figured it needed a blog post.
There are a bunch of tactical opportunities to measure quality in various processes, like the QA data you might extract from a production line for example. And while those details interest me, this thought process always bubbles up to the aggregate concept: what’s a consistent measure of quality across any product or service?
I have a short answer, but while you’re here I’ll walk you through how I get there. Including some examples of things I think are of high quality.
One of the reasons this question is interesting, is that it’s quite common to divide up data into quantitative and qualitative buckets. Often splitting the crisp metrics we use as our KPIs from the things we think indicate real quality. But, if you care about quality, and you operate at ‘scale’, you need a quantitative measure of quality.
On that note, in a small business or on a small project, the quality feedback loop is often direct to the people making design decisions that affect quality. You can look at the customers in your bakery and get a feel for the quality of your business and products. This is why small initiatives are sometimes immensely high in quality but then deteriorate as they attempt to replicate and scale what they do.
What I’m thinking about here is how to measure quality at scale.
Some things of quality, IMHO:
This axe is wonderful. As my office is also my workshop, this axe is usually near to hand. It will soon be hung on the wall. Not because I am preparing for the zombie apocalypse, but because it is both useful as a tool, and as a visual reminder about what it means to build quality products. If this ramble of mine isn’t enough of a distraction, watch Why Values are Important to understand how this axe relates to measures of quality especially in product design.
This toaster is also wonderful. We’ve had this toaster more than 10 years now, and it works perfectly. If it were to break, I can get the parts locally and service it myself (it’s deliberately built to last and be repaired). It was an expensive initial purchase, but works out cheap in the long run. If it broke today, I would fix it. If I couldn’t fix it for some extreme reason, I would buy the same toaster in a blink. It is a high quality product.
This is the espresso coffee I drink every day. Not the tin, it’s another brand that comes in a bag. It has been consistently good for a couple of years until the last two weeks when the grind has been finer than usual and it keeps blocking the machine. It was a high-quality product in my mind, until recently. I’ll let another batch pass through the supermarket shelves and try it again. Otherwise I’ll switch.
This spatula looks like a novelty product and typically I don’t think very much of novelty products in place of useful tools, but it’s actually a high quality product. It was a gift, and we use it a lot and it just works really well. If it went missing today, I’d want to get another one the same. Saying that, it’s surprisingly expensive for a spatula. I’ve only just looked at the price, as a result of writing this. I think I’d pay that price though.
All of those examples are relatively expensive products within their respective categories, but price is not the measure of quality, even if price sometimes correlates with quality. I’ll get on to this.
How about things of quality that are not expensive in this way?
What is quality music, or art, or literature to you? Is it something new you enjoy today? Or something you enjoyed several years ago? I personally think it’s the combination of those two things. And I posit that you can’t know the real quality of something until enough time has passed. Though ‘enough time’ varies by product.
Ten years ago, I thought all the music I listened to was of high quality. Re-listening today, I think some of it was high-quality. As an exercise, listen to some music you haven’t for a while, and think about which tracks you enjoy for the nostalgia and which you enjoy for the music itself.
In the past, we had to rely on sales as a measure of the popularity of music. But like price, sales doesn’t always relate to quality. Initial popularity indicates potential quality, but not quality in itself (or it indicates manipulation of the audience via effective marketing). Though there are debates around streaming music services and artist payment, we do now have data points about the ongoing value of music beyond the initial parting of listener from cash. I think this can do interesting things for the quality of music overall. And in particular that the future is bleak for album filler tracks when you’re paid per stream.
Another question I enjoy thinking about is why over the centuries, some art has lasting value, and other art doesn’t. But I think I’ve taken enough tangents for now.
So, to join this up.
My view is that quality is reflected by loyalty. And for most products and services, end-user loyalty is something you can measure and optimize for.
Loyalty comes from building things that both last, and continue to be used.
Every other measurable detail about quality adds up to that.
Reducing the defect rate of component X by 10% doesn’t matter unless it impacts on the end-user loyalty.
It’s harder to measure, but this is true even for things which are specifically designed not to last. In particular, “experiences”; a once-in-a-lifetime trip, a festival, a learning experience, etc, etc. If these experiences are of high quality, the memory lasts and you re-live them and re-use them many times over. You tell stories of the experience and you refer your friends. You are loyal to the experience.
Bringing this back to work.
For MoFo colleagues reading this, our organization goals this year already point us towards Quality. We use the industry term ‘Retention’. We have targets for Retention Rates and Ongoing Teaching Activity (i.e. retained teachers). And while the word ‘retention’ sounds a bit cold and business like, it’s really the same thing as measuring ‘loyalty’. I like the word loyalty but people have different views about it (in particular whether it’s earned or expected).
This overarching theme also aligns nicely with the overall Mozilla goal of increasing the ‘number of long term relationships’ we hold with our users.
Language is interesting though. Thinking about a ‘20% user loyalty rate’ 7 days after sign-up focuses my mind slightly differently than a ‘20% retention rate’. ‘Retention’ can sound a bit too much like ‘detention’, which might explain why so many businesses strive for consumer ‘lock-in’ as part of their business model.
Talking to OpenMatt about this recently he put a better MoFo frame on it than loyalty; Retention is a measure of how much people love what we’re doing. When we set goals for increasing retention rate, we are committing to building things people love so much that they keep coming back for more.
- You can measure quality by measuring loyalty
- I’m happy retention rates are one of our KPIs this year
My next post will look more specifically about the numbers and how retention rates factor into product growth.
And I’ll try not to make it another essay. 😉
It was a long weekend here in the UK which gave me some time to make some final adjustments to a desk I’ve been building and take a photo. And, here it is.
Many years ago I read Douglas Adams’ Last Chance to See (which on a side note was one of my original motivators for wanting to work at WWF). Though I was maybe fifteen when I read this book, one of several notions that stuck with me was:
In the twenty years following, airports have been upping their game when it comes to design. Standing Desks however are severely lacking in the ‘pretty’ department (even if they’re currently trendy). If you haven’t looked into this yourself, try an image search now for ‘Standing Desk’. Even try a Pinterest search where you limit the results to the things that design conscious human beings think are worth looking at, and you’ll see standing desks are struggling.
There are a few exceptions, including the craftsmanship of MoFo’s own Simon Wex. But mostly, standing desks are not very pretty.
But still, I wanted a standing desk, and one that I’d want to look at for the many many of the hours in my life spent working. And I didn’t feel like spending hundreds of pounds buying an ugly office version, so I challenged myself to build something not-ugly. For relatively little money.
To credit my sources, I basically combined the two desks below into the thing I wanted (and added an extra shelf to stash the laptop):
- This lovely desk
- The approximate angles and measurements of this standing desk (in particular the negative slope keyboard stand)
The three horizontal shelves were IKEA shelves left in our house by the previous owners, so I spent about £25 on the remaining timber and carriage bolts to hold the frame together.
I’ve been using it for a few weeks now, and am thoroughly enjoying it.
Also, making things is fun.
I wrote a version of this strategy in January but hadn’t published it as I was trying to remove those ‘ish‘s from the title. But the ‘ish’ is actually a big part of my day-to-day work, so this version embraces the ‘ish’.
MoFo Metrics Measures of Success:
These are ironically, more qualitative than quantitative.
- Every contributor (paid or volunteer) knows at any given time what number they (or we) are trying to move, where that number is right now, and how they hope to influence it.
- We consider metrics (i.e. measures of success) before, during and after after each project.
- We articulate the stories behind the metrics we aim for, so their relevance isn’t lost in the numbers.
- A/B style testing practice has a significant impact on the performance of our ‘mass audience’ products and campaigns.
1. Every contributor (paid or volunteer) knows at any given time what number they (or we) are trying to move, where that number is right now, and how they hope to influence it.
- “Every” is ambitious, but it sets the right tone.
- This includes:
- Public dashboards, like those at https://metrics.webmaker.org
- Updates and storytelling throughout the year
- Building feedback loops between the process, the work and the results (the impact)
2. We consider metrics (i.e. measures of success) before, during and after after each piece of work.
- This requires close integration into our organizational planning process
- This work is underway, but it will take time (and many repetitions) before it becomes habit
3. We articulate the stories behind the metrics we aim for, so their relevance isn’t lost in the numbers.
- The numbers should be for navigation, rather than fuel
4. A/B style testing practice has a significant impact on the performance of our ‘mass audience’ products and campaigns.
- This is the growth hacking part of the plan
- We’ve had some successes (e.g. Webmaker and Fundraising)
- This needs to become a continuous process
Those are my goals.
In many cases, the ultimate measure of success is when this work is done by the team rather than by me for the team.
We’re working on Process AND Culture
Process and culture feed off of and influence each other. Processes must suit the culture being cultivated. A data driven culture can blinker creativity – it doesn’t have to, but it can. And a culture that doesn’t care for data, won’t care for processes related to data. This strategy aims to balance the needs of both.
I tried to write one, but basically this strategy will respond to the roadmaps of each of the MoFo teams.
So, what does Metrics work look like in 2015?
- Building the tools and dashboards to provide the organisational visibility we need for our KPIs
- ‘Instrumenting’ our products so that we can accurately measure how they are being used
- Running Optimization experiments against high profile campaigns
- Running training and support for Google Analytics, Optimizely, and other tools
- Running project level reporting and analysis to support iterative development
- Consulting to the Community Development Team to plan experimental initiatives
Plus: supporting teams to implement our data practices, and of course, the unknown unknowns.
I’ve blogged about various experiences of online learning I’ve taken part in over the years and wanted to reflect on the most recent one. Coursera’s three week Introduction to Ableton Live.
Learning more about learning is one of my personal goals this year. And I find writing out loud to be useful tool in thinking. So that’s mostly the point of this.
I take these courses mostly because I like learning new things, but also because I’m interested in online learning more generally. How do you most effectively transfer knowledge, skills and motivation via the web, and/or about the web? That question is often on my mind.
Almost all of the projects I work on at Mozilla are somewhere in the education space; directly with Webmaker or Mozilla Learning Networks and tangentially in the topic of volunteer contribution. Contributing to an open source project as complex and distributed as Mozilla is a learning experience in itself, and sometimes requires specific training to even make it possible.
To further frame this particular brain dump, I’m also interested generally in the economics of the web and how this shapes user experiences, and I have strong feelings about the impact of advertising’s underlying messaging and what this does over-time when it dominates a person’s daily content intake. I’m generally wary of the word “Free”. This all gets complex when you work on the web, and even directly on advertising at times. Most of my paycheques have had some pretty direct link to the advertising world, except maybe when I was serving school dinners to very rich children – but that wasn’t my favourite job, despite it’s lack of direct societal quandaries.
Now, to the content…
If you’re like me, you will tend to read notes about a topic like ‘commerce in education’ and react negatively to some of these observations because there are many cases where those two things should be kept as far apart as possible. But I’m actually not trying to say anything negative here. These are just observations.
All roads lead to… $
My online experience within the Coursera site was regularly interrupted with a modal (think popup) screen asking if I wanted to pay to enrol in the ‘Signature Track’, and get a more official certification. This is Coursera’s business model and understandably their interest. It wasn’t at all relevant to me in my life situation, as I was taking a course about how to play with fun music software in my free time. I don’t often check my own qualifications before I let myself hobby. Not that anyone checked my qualifications before they let me work either, but I digress. Coursera’s tagline says ‘free’, but they want you to pay.
All assignments for the course had to be published to Blend for peer-evalutation, Blend is like Github but for raw audio production tracks rather than source-code. I didn’t know about Blend before the course, and really like it as a concept and how it’s executed and for what it could do for collaborative music making. But I note, it is a business. This course funnels tens of thousands of new users into that business over the course of a few days. There might not be any direct financial trade here (between companies for example), but users are capital in start-up land. And I now receive emails from Blend with advertisements for commercial audio production tools. My eyeballs, like yours, have a value.
While hosted on Coursera, the content of this course is by Berklee College of Music. The content they ‘give away’ would traditionally only have been available to paying students. Berklee’s business is selling seats in classes. This course isn’t given away as an act of kindness, it’s marketing. Three weeks is short and therefore the content is ‘light’. Lighter than I was expecting (not that I’m entitled). But halfway through, we receive a promotional email about Berklee’s own online education platform where you could create an account to get access to further ‘free’ videos to supplement the Coursera materials. I found these supplementary videos more useful, and they lead to offers to sign-up for extended paid courses with Berklee Online. For Berklee, this whole excercise is a marketing funnel. Quite possibly it’s the most fun and least offensive marketing funnel you can be dropped into, but it exists to do that job.
Now, I write this with genuine sympathy, as I’ve walked the floor at countless venues trying to sell enough music and merch to cover the petrol costs of playing a gig. But this is a commercial element of this learning experience, so I will note it. At many points throughout the three weeks, we had opportunities to buy Erin’s music, t-shirts, and audio production stems (these are like a layer file of an original recording) for consumption and or remixing. I know you have to hustle if you’re making music for a living, but the observation here is that the students of this course are also a marketable audience. Perhaps only because they arrive en-mass and end up slightly faceless. I’m sure it would be weird for most teachers to sell t-shirts in a class-room. It wasn’t particularly weird online, where we’re desensitised to being constantly sold things. And I may have only noticed this because I’m interested in how all these things fit together.
The course was about learning Ableton Live. A commercial audio production tool. So at some point, the cost of Ableton had to be considered. Ableton offers a free 30 day trial, which works for this course and they kindly (or sensibly) agreed to let people taking the course start a new trial even if they’d used their 30 days already. Good manners like those are good for business. Anyway, I already owned Live 9 Intro (aka the cheap version), and for a three week intro course it does more than enough to learn the basics (I guess that’s why it’s called Intro?). But the course taught and encouraged the use of Live 9 Suite (the EUR599 rather than the EUR79 version). Until some people complained, the use of features in Suite was required to complete the final assignment. Reading between the lines, I doubt there was any deliberate commercial discussion around this planning, but the planning definitely didn’t stem from the question: ‘how can we keep the cost down for these beginners?’. At the end of the course there were discount codes to get 15% off purchasing anything from Ableton. I didn’t use Suite during the course, but I’m playing with it now on my own time and terms, and may end up spending money on it soon.
It’s wonderful, but it’s not Wikipedia. The course opened a lot of doors, but mostly into places where I could spend money, which I am cautious about as a model for learning. It was valuable to me and prompted me to learn more about Ableton Live than I would have done in those three weeks without it. So I’m grateful for it. But I can’t in my heart think of this as a ‘shared public resource’.
For my own learning, I like deadlines. Preferably arbitrary. The fact that these Coursera courses are only available at certain times during the year, really works for me. But I struggle with the logic of this when I think about how best to provide learning material online to as many people as possible. The only MOOC style courses I have finished have been time-bound. I don’t know how many people this is true for though.
People will learn X to earn Y. For me this course was a form of hobby or entertainment, but much learning has a direct commercial interest for students as well as educators. Whether it’s for professional skills development, or building some perceived CV value.
There is no ‘free’ education, even if it says “free” on the homepage. There is always a cost, financial or otherwise. Sometimes the cost is borne by the educator, and sometimes the student. Both models have a place, but I get uncomfortable when one tries to look like the other. And if the world could only have one of these models for all of education I know which one I’d choose. Marketing fills enough of our daily content and claims enough brainprint as it is.
I thought I might find some conclusions in writing this, but that doesn’t always happen. There are a lot of interesting threads here.
So instead of a conclusion, you can have the song I submitted for my course assignment. It was fun to make. And I have this free-but-not-free course to thank for getting it done.
First, I’ll note that even taking the time to write these short ‘note to self’ type blog posts each week takes time and is harder to do than I expected. Like so many priorities, the long term important things often battle with the short term urgent things. And that’s in a culture where working open is more than just acceptable, it’s encouraged.
Anyway, I have some time this morning sitting in an airport to write this, and I have some time on a plane to catch up on some other reading and writing that hasn’t made it to the top of the todo list for a few weeks. I may even get to post a blog post or two in the near future.
This week, I have face-to-face time with lots of colleagues in Toronto. Which means a combination of planning, meetings, running some training sessions, and working on tasks where timezone parity is helpful. It’s also the design team work week, and though I’m too far gone from design work to contribute anything pretty, I’m looking forward to seeing their work and getting glimpses of the future Webmaker product. Most importantly maybe, for a week like this, I expect unexpected opportunities to arise.
One of my objectives this week is working with Ops to decide where my time is best spent this year to have the most impact, and to set my goals for the year. That will get closer to a metrics strategy this year to improve on last years ‘reactive’ style of work.
If you’re following along for the exciting stories of my shed>to>office upgrades. I don’t have much to show today, but I’m building a new desk next and insulation is my new favourite thing. This photo shows the visible difference in heat loss after fitting my first square of insulation material to the roof.
What’s happening this week?
My number one goal (P1) for this week is solving offline friendly mobile analytics for Webmaker App, while keeping other projects ticking along adequately.
Here’s to a productive week.
I don’t want these weeknotes to be a complete ‘done list’, as we have enough of those internally. This just a quick reflection on the extra objectives set for the week.
- I moved outside into the MVP garden office (it’s still looking mostly like a shed).
- This was thanks to the magical powers of powerline adapters which I only recently heard about. I still do not understand the sorcery that is transferring a high speed network through the existing electrical circuit, but it’s working without me needing to run any cabling.
- I spent enough time chipping away at my processes and on ‘working open’ that I’m feeling good about it, and still enough time getting things done.
- I cleaned out my ‘mofo-metrics’ Bugzilla backlog from 2014, and killed lots of tickets that weren’t relevant any more.
- Setup my first Data Practices Review ticket as part of the new ‘Data Steward’ work I have acquired this year.
I feel like I’m getting in to the flow of 2015 a bit now.
One of this week’s conversations was with Nesta, about Webmaker usage within the UK and whether or not we have data to support the theory that face-t0-face events have an impact getting people involved in making on the web. These are two topics that interest me greatly.
I’m basically copying some of my notes into blog form so that the conversation isn’t confined to a few in-boxes.
And the TL;DR is our data represents what we’ve done, rather than any universal truth.
Our current data would support the hypothesis that face-to-face time is important for learning, but that would simply be because that’s how our program has been designed to date. In other words, our Webmaker tools were designed primarily for use in face-to-face events, which meant that adoption by ‘self-learners’ online is low because their is little guidance or motivation to play with our tools on your own. This year we’re making a stronger push on developing tools that can be used remotely, alongside our work on volunteer led face-to-face events. This will lead to a less biased overall data set in the future where we can begin to properly explore the impact on making and learning for people who do or don’t attend face-to-face events at various stages in their learning experience. In particular I’m keen to understand what factors help people transition from learners, to mentoring and supporting their peers.
I also took a quick look at the aggregate Google Analytics location data for the UK audience which I hadn’t done before and which re-enforces the point above.
Above: Traffic to Webmaker (loosely indicating an interest in the topic) is roughly distributed like a population map of the UK. This is what I expect to see of most location data.
Above: However, if you look at the locations of visitors who make something, there are lots of clusters around the UK and London is equaled by many other cities.
To-date, usage of the Webmaker tools has been driven by those who are using the tools to teach the web (i.e. Webmaker Mentors). But we also know there are large numbers of people who find Webmaker outside of the face-to-face event scenarios who need a better route into Webmaker’s offering.
The good news is that this year’s plans look after both sets of potential learners.