Product Bugle 8/7: Prioritising for growth

What's a 'Strategic Priority'? How do you make sure they happen? How they built up Amazon's Video business, and what it means for Netflix. My last newsletter before a summer break.

Photo by Bill Jelen on Unsplash

A couple of weeks ago, I wrote up my recipe for how to create a product strategy - bringing together a lot of (other people’s!) ideas into a single recipe.

This week, I wanted to write about prioritisation. But as part of it, I’m going back to thinking about strategy. Because prioritisation is all about making sure that the most important stuff happens first, and to do that, you’ve got to know what the most important stuff is.

This is long. Apologies. Think of it as a parting gift before I head off on my holidays for a couple of weeks.

Not all prioritisation is the same

Most of the time when I read about prioritisation the normal reference point is some kind of scoring framework (more of them later) either a relatively lightweight ‘Impact vs Effort’ matrix; a full blown RICE scoring system or the KANO model, which tbh, I’ve never found practically useful (I’m willing to accept that’s me).

Now if you’re sitting one a nice tidy backlog of relatively similar sized items, with no internal or external pressure and you’re working what to put in your next, I can see how you might be able to score your way to decision making.

However, while I’ve been writing this, I’ve had a number of conversations with CEOs of decent sized start ups; and product leaders in corporate teams all saying the same thing: we know we need to build - we’re just not clear what we need to focus on..where we need to start.

This is a bigger, higher level problem. And it’s where prioritization and strategy intersect. So, I’m going to focus on this: Strategic priorities. And I’m going to talk a lot about video - a sector I know well - and a lot of big B2C companies. But the same principles apply regardless of scale or sector.

Three degrees of ‘prioritisation’

One of the problems is that ‘prioritisation’ is often addressed as a single thing - whereas I’ve found there are really three distinct processes, and in each case the people, information and processes required are different.

1. Strategic prioritisation: Deciding on the major growth-drivers for your product and business.

2. Sequencing of delivery: Working out which order things are actually going to get built in - which can often be influenced by both commercial and technical dependencies or contractual/ regulatory urgency.

3. Micro prioritisation: The close up prioritisation that happens when you’re actually in or very close to the build process and you have to make decisions about what ‘sub-features’ within a feature you will actually ship.

A lot of what’s written about prioritisation focusses heavily on steps 2 and 3 - which is where scoring frameworks can be helpful. This week - I’m just going to focus on the first item, because this is where the most important decisions need to be made.

A lot of the examples I’m going to give here are big, consumer facing businesses that had hundreds or people working on them. The same though is just as true

Strategic priority type 1: driving growth through reappraisal.

If you’re lucky, your product is so perfect that you are going to achieve stellar growth just through a series of tactical optimisations and A/B tests. If so - feel free to optimise away and skip this. But, as you can see in my sidebar about Netflix below, that growth might not last forever.

However, real growth - as in step-change, 10X growth: the kind that CEOs and investors crave, and CPOs dream of having on their CVs; rarely comes from incremental improvements.

More likely it’s the result of significant changes that either get your existing customers to use your product in new ways or encourage new customers to use your product. In both cases you are driving a reappraisal of your product and driving growth of the back of it.

Going back to my post about strategy - this is all about ‘Where you’re going to play; and how you’re going to win’.

Think of adding the marketplace to Facebook; Netflix’s international expansion; or the way that LinkedIn adding Jobs, and then it’s ‘article’/ ‘blogging’ functionality; or Spotify adding podcasts and, soon, Audiobooks; Google adding Shorts to YouTube; or Instagram launching Stories. Or to give to recent examples from my last role: Sky adding Netflix and Amazon Prime to its Sky Q set top box or launching Sky Glass so you don’t need a satellite dish.

As you can see, sometimes these strategic moves aren’t actually that innovative. In most cases these are feature or proposition changes that have been done before. Sometimes they’re shamelessly defensive to try and neutralise new competition. YouTube Shorts to try and emulate TikTok; Facebook marketplace was a counter to eBay; or Instagram Stories emulating Snapchat.

This is a reappraisal: ‘Hey, that thing you were heading over there for…you can now do it over here’.

But what you’re trying to do is bolt on more things that are - or are about to hit the orange bit of this graph (stolen from here) - effectively strapping a rocket pack onto your business.

The fact these things aren’t ‘innovative’ - ie totally new ideas doesn’t make them easy. Working out where Shorts will fit into YouTube; or building a new backend to support Audiobooks on Spotify - and then how it will fit in with the existing UI is an act of heavy lifting all the way through a complex software stack. The ideas might not be ‘new’ but there’s plenty of innovation required to make them happen.

As I’ll talk about a bit when it comes to prioritisation. You are at best going to launch with half a product (but never a half-assed product!). And it might well take a fair bit of refinement to get it right. ( I think Amazon had three goes to get it’s Third Party Sales/ Marketplace right).

Strategic priority type 2: growth enablers

Now all of the above are quite visible and tangible growth drivers. But there are also back end drivers of growth: platforms and capabilities that can scale to both enable and support growth.

This could be: bringing together two platforms that have co-existed with a company after a merger; moving from a big ball of code to a service driven architecture; developing a new data platform to be able to get insights within minutes rather than days (or not at all); shifting to a new payment provider to enable global expansion.

These

Strategic by name, tricky by nature

These kind of initiatives all have similar characteristics

• Even in their most minimal form, these will be big. The default position will be ‘high impact, high effort’.

• They often require multiple teams in order to be delivered

• They are very often difficult to test in any meaningful way. You can research; you can possibly user test a prototype, but at some point you have to make a big commitment to build.

• Even when you have chosen your strategic priorities - you will need to prioritise aggressively within them to agree on a scope for launch. Assume that you will at best, be able to launch half of what you could launch in a reasonable time frame.

The point is that these kind of initiatives don’t happen by accident; and even when you’ve ‘launched’ you still have a ton of work to do after. Here’s an example of them in action..

Amazon Prime Video - one strategic priority after another

The history of Amazon Prime Video, where I worked for 5 years, has been a succession of ‘strategic’ initiatives that have repeatedly got customers to reappraise the product - bringing in new customers and getting existing customers to find new purposes for it.

Some of these (not all delivered in my time), and the reappraisals they triggered were..

• The launch of Amazon Video - a transactional video on demand service (TVOD).
  Reappraisal: Oooh..Amazon I can now buy stuff and stream it or download it and not wait for a DVD.

• The launch of Prime Video.
  Reappraisal: I can now stream video as part of my Prime membership.

• Distribution onto PlayStation, Xbox, Roku, LG, Samsung TVs, iOS, Android
  Reappraisal: Now I can get it on that device in my living room and wherever I am

• The launch of Prime Video in UK and DE:
  Reappraisal..and where I live..! (if you’re British of German)

• The launch of Amazon Channels.
  Reappriasal: I can now get Starz and a host of other channels all in one place.

• Going global:
  Reappraisal: I can now get Amazon video anywhere in the world

• Launching live sport
  Reappraisal: NFL, Tennis, Premier league football - making it now relevant for tens of millions of sports fans.

And powering this, was a parallel series of platform work to support streaming, catalogue management, content discovery, account management and payment.

The point about all these things is that with each one is that they open up new customers and new use cases for existing customers.

The other point is that when say Amazon Channels was being launched, or when we were launching Prime in the UK and DE, it was almost impossible for anything else to happen. All resource had to be primarily focussed on this. I say ‘almost impossible’ - because other things happened, but never at the risk of slowing down our top priority.

Now, this wasn’t always fun. There were features I wanted to launch and neat A/B tests I wanted to run. But it was right. This is what prioritisation is all about: tough calls.

We’d learned from past launches is that new features per se, hardly ever drove growth. but distribution - geographically and onto new devices did. Adding in new content types did. A new feature might improve some engagement metric and have an overall positive impact, but when you’re planning.

The other thing to stress it that all of these things were brutally prioritised for launch. The first iOS app, for example, didn’t have search. And once they were live they then took on a life of their own - they too had backlogs; they needed to be nurtured and optimised. They added to the overall workload.

A sidebar about Netflix

I love Netflix. And working in video for a decade, I can’t think how many times I’ve seen a feature on Netflix that I wish we had (and then had to explain to an exec why we can’t build it ‘yet’).

Part of Netflix’s genius is that they have focussed on a single model: Flat subscription. This is part of what makes the user experience so great. You pay once and you can use everything you see on the service. It makes everything simpler.

And since the move from DVDs to streaming their ‘product-driven’ growth has come from a few big hitting ‘strategic priorities’: global expansion; and ubiquitous distribution (including those genius buttons on remote controls!) and partnerships (such as the one with Sky). And, I’d argue, one big feature: ‘Profiles’. All of these drove the kind of reappraisal needed for growth. With profiles it was: ‘Oh this is for me, as well as my kids..’

But those cards have now been played. And what they’re left with is the much harder task of content driven growth: the next big movie or TV show. It’s true that when they launched ‘The Crown’ and started doing award winning movies straight on their service that drove some reappraisal. But that’s a slog.

What we’ve been experiencing is a brilliantly tuned optimisation machine. One A/B test after another they get better and better at getting us to our content; and using data to optimise their content selection. But at their scale, none of this really drives the kind of reappraisal-driven growth they need to show. It leaves them much more subject to the natural product lifecycle, especially when they have Disney+; Amazon, Apple and others all getting better.

So now what? Well the next two things are adding advertising (allowing a possible reappraisal if it means there’s a much cheaper package available as a result); and clamping down on password sharing (a kind of negative reappraisal - ‘shit..now I need to pay for it instead of using my cousin’s password’).

I’ve seen Reed Hastings talk in the past about not moving into sport. His argument was that the only thing streaming bought to sport was delays and buffering - but that’s moved on. I suspect the content costs might not work for them either (you pay a lot, and it’s all gone the second the event finishes). But, over the next 3 - 5 years, I can’t help thinking they’re going to have to break into a new category, even if it means sacrificing a bit of simplicity, if they want to push back against all that new competition.

Weren’t we meant to be talking about prioritisation?

Ah yes - sorry about that. Look you’re smart. And if you’ve been motivated enough to get this far, I’m sure you’ll now have worked this out yourself.

Here’s a recap on the logic so far.

• Real growth is the outcome of substantial changes to your product. These are normally big, complex initiatives that will require a big chunk of your total resource (no matter how big your organisation) to be engaged in.

• If you want to make them happen - they have to be prioritised often at the expense of other items which will look like easier quicker wins.Like any personal prioritisation system - you have to start by sorting out the most important stuff; and let everything else build around that.

• You will have to ‘prioritise within your priority’ - assuming you’re going to launch ‘half a product’, there’s a lot of work to be done to work out exactly what that half a product is.

• Once it’s live - it will create more work if you really want to get it right. In other words, you’re creating a bigger prioritisation challenge for the future.

So - first things first. Scrap your simple frameworks and scoring systems. Build your list of your top 3 growth drives (which is obviously a collaborative effort between product, business teams, engineering and design), and then you have to prioritise around them.

You know the whole rocks, pebbles, grain of sand analogy don’t you? Don’t sweat about the grains of sand, until the rocks are in place. Some possible secenarios

• You take on one first, and so allow other work to happen in parallel.

• You take on one or more item have to block off everything except essential maintenance.

• You focus on a large ‘enabler’ that a platform team can work on independently - but it just means that everyone else will now have to reprioritise to remove dependencies on that team.

• There’s some smart way you can allow some teams to continue to work independently on other stuff without slowing this down (as Sky we launched Disney+ and a series of enhancements to Sky Q while we were working on SkyGlass; at Amazon we launched HDR on the service while we were working on Amazon Channels).

There is no solution here where everyone gets everything. It can be simple (right we are now all going to focus on ‘X’!). But it’s never easy. These are the kind of tough calls that are part of the real world of product and engineering leadership. But the point is: get the big stuff sorted and let the small stuff slot into place around it.

In the short term, this is always painful. Lots of incremental improvements get delayed or never happen at all. But, in the long term, it pays off, big time.

TV this week..

Well, it’s got to be this…the annual post Casa Amor meltdown.

And now, off for the summer..

I’m taking a break from the Bugle until the end of July. I’m Looking at a few options for some changes / enhancements in the Autumn. As always - any feedback, or suggestions for things to cover will be most welcome.