Lighthouse thinking® and the path to extraordinary outcomes. Transformation Lessons from the Amazon Playbook 

In this article we are going to explore the power of lighthouse thinking. Lighthouse thinking  is an alternative approach to innovation and transformation. Its power lies in two critical factors that differentiate it from the crowded field of transformation and innovation theory.

The first is it de-risks significant change, a factor that stems from the second and which is also critically important given the high rate of failures: it inherently creates a learning environment directed at the future operating model or key innovations. Why is this important?

We often do not know what the outcome of a transformation should be, in any tangible sense, and innovations contain many unknowns. A company needs to define its future-state iteratively as the knowledge and skills to understand the future grow. 

Very few companies are like Amazon in this respect. Because of its use of Lighthouse Projects, Amazon has industrialised innovation and operating model change. 

In two years between 2018-2020, it launched at least five new platform businesses while the rest of the business world puzzled over what a platform business model really is. Overall, it now manages at least twenty thematic business platforms at the same time as innovating in AI, Cloud services and applications, voice interaction in the home, global logistics, finance and video content.

In ecommerce, the pace of change is made possible by the lighthouse. Lighthouse thinking is a rescue remedy that creates the know-how to understand and define whatever changes might be necessary to get to extraordinary new outcomes.

Amazon is a master practitioner of lighthouse projects and by using its recent history as a case study:

  • we can start to illuminate key factors in its unprecedented success 
  • and at the same time shape our thinking about how the lighthouse approach should work.

Those are the key lessons of this article. We will draw them out by using Amazon’s Explore platform as an example. 

Explore is a livestream platform for ecommerce, the first of its kind from Amazon, though the technique is now established in China. 

Amazon’s approach is to use Explore as a lighthouse project. Going beyond simply being an innovation or a pilot, these projects allow large companies to remodel their activities quickly in response to opportunities or threats. 

We will discuss below how Amazon routinely uses this technique to adapt.

Agile ways of work tend to stop at minor adaptations to work practices. Having a technique to be strategically agile is the holy grail of business transformation. The lighthouse fits the bill because it shifts the emphasis in transformation away from large, marque programs to simpler, low risk but easily executed and dynamic ones that embrace powerful learning experiences.

Amazon and the lighthouse imperative

Amazon is the ecommerce incumbent every challenger needs to beat. 

Notwithstanding the fact that shock factors such as pandemic or energy price volatility create significant economic shifts, generally speaking, technology is what creates significant competitive shifts.

Nimble competitors tend to rely on new technologies to tease out an opportunity to make inroads into an incumbents customer base.

Such a shift is taking place around consumer generated livestreaming. It may seem a niche activity right now but in China, livestreams are becoming the default content for ecommerce sales, prompting Forbes to describe it as  “the rage”.

Livestreams shift retailing from static databases of content (product pages, SKUs, customer accounts, reviews) to a contrasting landscape of dynamic content-creation by empowered users.

Amazon’s operating model lies with the former, whereas fast growing companies, including the brand of the moment, TikTok, have pioneered the latter (and indeed the online magazine Quartz has described TikTok as a sleeping ecommerce giant) . 

However, Amazon is in a paradoxical position with livestreaming. It owns Twitch (now part of Prime gaming), the premier livestream for games.  Yet, in ecommerce, rather than graft on lessons from Twitch, it is attempting to learn its way to relevance through new lighthouse projects.

Amazon, of course, has a track record of creating change its own way and one technique it has used consistently over time is the lighthouse. 

To stimulate a broader discussion about incumbent responses and learning strategies through a lighthouse we will:

  • Introduce lighthouse thinking
  • Clarify the idea of a lighthouse project
  • Briefly describe how Amazon routinely deploys lighthouse projects
  • Explain how the Explore platform works
  • And discuss the ways that Explore impacts Amazon’s work practices and its pursuit of new business opportunities

 

Lighthouse thinking, some background

In our book Transformation Sprint, we spelled out why transformation leaders should design change around Lighthouse Projects rather than big-bang transformations.

The reason for this, on the negative side, is that big projects fail too often and the evidence is now cumulative that they just don’t get enterprises to where they need to be. Big-bang change fits the mode of consultant advisors rather than the needs of the enterprise client. But transformation itself is widely misunderstood, being seen mostly as a singular transition from an AS IS history to a TO BE (or target operating model) future.

On the positive side, lighthouse thinking gives you projects that are designed to have a structural impact, that is to enable you to change, but through small and low risk steps. This is the most obvious characteristic and there are, in fact, many other advantages that we will spell out.

Targeted project design

Analysis of the kind we do in transformation sprints might lead you to conclude that your priority problem is a lack of genuinely customer-centric innovation and that you need to skill up on discovering new value. In effect, you need to improve your innovation model. That might lead you to a lighthouse project around deeper customer insight and data, all primed for more rapid, iterative delivery.

It could equally be that specific processes are blocking all kinds of progress. That might, typically, be incompatible core platforms that have become over complex. The priority has to be to resolve these because of the volume of dysfunction they create. These lead to the kind of lighthouse project that Amazon used in the 2000s when it developed Elastic Compute Cloud, the precursor of AWS.

Or maybe the progress-blocker is the portfolio of work. Too many projects and not enough relevance. If so, then a value management review could be the lighthouse that allows you to strip out wasted resources and refocus on value.

New ways to work

Lighthouse projects are designed to address structural blockers :

  • value discovery (the innovation model), 
  • value management, 
  • value delivery or more broadly, skills and process model innovations. 

They are also designed to do this through new ways to work with a strong emphasis on creating a strong learning environment, animated by visual work techniques (so called working out loud).

These small, iterative steps can be designed to have extraordinary outcomes. Their power and relevance comes from the way they cut through the noise around transformation, simplifying the unnecessarily complex.

The average enterprise is already dealing with several transformations at any one time, so de facto wrestles with piecemeal changes. These can be, dealing with the introduction of AI/ML, the rise of IoT, an incomplete digitisation of processes, the introduction of agile ways to work, culture change, highly distributed location (remote work) and so on.

Problems with multiple change-tracks begin when they are lumped into one big program with a big budget allocation, a transformation director, and a target operating model.

We have argued several times now that this approach shapes the transformation in the wrong way. What’s needed is an agile approach to transformation. This is best done through a generative operating model, driven by a lighthouse design.

Lighthouse projects and generative or adaptive operating models

So what do these important terms mean? 

A generative operating model is one where there is no fixed endpoint to a transformation. That does not mean it has no targets or goals. But it does mean we accept that the endpoint is always likely to change as a transformation teams gain experience.

If that is so, then why not set off with the goal of learning uppermost in our minds. Let’s design the “target” as the creation of a bone fide learning environment where we gain enough experience to validate what needs to change, grow the culture where we can develop consensus on priorities, and learn the skills both to do the change and execute successfully in the new environment.

The vehicle for that change is the Lighthouse Project, a relatively small project that has a structural impact, opens up new business possibilities and uses new, agile ways to work.

Amazon does this routinely through a paper exercise that involves writing the press release and the FAQ for a new service, a technique that forces staff to think through in detail how the finished service will benefit customers. 

This technique gave rise to AWS, one of the most successful business initiatives of all time (in the process allowing Amazon to solve many of its core platform problems). Having this dual feature of solving problems and creating new opportunities is also the hallmark of a good lighthouse.

Lighthouse Thinking embraces small, strategically significant projects that can play out both as catalysts for internal process changes (transformation) and as new business models.

The lighthouse project defined

In the 2018-2020 period, Amazon launched a series of new lighthouse projects seemingly designed to update its ecommerce environment with new customer interactions.

Amazon’s use of lighthouse projects, as distinct from mixing in staff from Twitch, is interesting for a number of reasons and not least among them is its recognition that staff have to learn through experience, a key benefit of a lighthouse.

Before going into more detail, let’s distinguish between lighthouse projects and other types of innovations such as: 

  • prototypes,
  • proofs of concept, 
  • MVPs 
  • and pilots.

If you need to demonstrate that a technology or component works, say in a different environment from the norm, then you need a proof of concept. An MVP is a good way to tease out the commercial potential of an idea. Pilots are good for assessing the resource needs (cash, but also personnel, skills and capabilities, access to distribution and so on) of a full rollout. A prototype is a good project format for showing that technological components can be assembled in a usable way.

But none of these give you access to the knowledge you need if you are to make a success of transformation. Prototypes and proofs of concept are relevant to a new service or product that might form part of a new direction but they don’t tell you a whole lot about new operating model needs. MVPs are accounting mechanisms that can give you incremental insight into what revenue levels might look like for any particular market context. It is not a fulcrum for a transformation.

In a transformation context, there’s no point in piloting a technology or a service unless it is tied back to understanding all the elements of change that a significant innovation might entail: new work processes, innovations in technological infrastructure, security implications, the role of the new service in a portfolio and the constraints it exercises on other resources, the need for partnerships and so on. But that is exactly what a Lighthouse gives. It is designed to yield the right collective experiences to grow that kind of knowledge.

And plus, it is really about being open-ended about that. If the target operating model is set in stone, there can be no lighthouse project. Lighthouse projects can focus on any aspect of your OM and also the processes, methodologies or even culture the underpinning it. The operating model has to be adaptive (or generative) enough that learning along the way, shapes future direction.

How Amazon deploys lighthouse projects

In Flow Magazine recently, we noted that Amazon’s Cloud services AWS, began as a lighthouse project. It arose from internal discussions about tangled and duplicated infrastructure, which became the subject of a thought piece on how infrastructure and applications might be separated. That paper allowed people intelligent speculation about different forms IT infrastructure could take. As it was, Amazon had replicated IT platforms across different departments and countries and had equally heterogeneous communications standards.

While the first step in creating something simpler was to introduce IP communications, the discussion kept coming back to not just standardising IT but changing how the company thought about infrastructure and then potentially using that as an asset to sell. A small, two person team got working on Elastic Cloud Compute, which ultimately became the bedrock for AWS.

Several points are important to note about this. Cloud services changed the way the company approached all work. The concept allowed Amazon to become a startup again, firing up new projects on elastic computing resources and easily available applications. But the project itself began and functioned in an extremely agile way, away from Seattle. It was an autonomous team working, And its impact was extremely strategic with a structural impact that reshaped infrastructure, culture and the future.

Amazon has done something similar with its employee health services program, though that program is bound to be less dramatic in its impact. 

The company had ideas about improving employee healthcare at its Seattle headquarters, in particular, making greater use of remote facilities (long before COVID struck). Partnering with Berkshire Hathaway and Microsoft, Amazon implemented a new service that gave employees access to remote diagnostic and psychological support. The rollout to two other companies allowed Amazon to test how these services could work from various standpoints: Infrastructure, confidentiality, incentives for use, impact on productivity and so on.

With its Spark platform Amazon tried to develop a collaborative shopping discovery environment. This is how the Verge describes Spark: “It prompted customers to pick a selection of interests in the section in the feature, and would then show you a feed of posts from users that related to those interests. The service seemed to be designed to replicate influencers using social networks like Instagram and Pinterest. Users could react to posts with a “smile” or a comment.”

Spark lasted a total of two years, launching in 2017 and closing in 2019. The 2017 launch came after a beta period, so clearly Amazon felt it had the skills to make Spark work. 

The service attempted to create visually anchored social networks in the way Instagram does. 

In Spark, it is possible to see how Amazon has institutionalised Lighthouse projects around platform technology. They are quick to fire up and equally quick to drop and forget their lighthouse projects, which is as it should be. 

As the company wound Spark down, it began working on its Live and Explore platforms. 

To underline the point further, in the same time period. it was also developing and launching Amazon Handmade (an Etsy lookalike), Amazon Subscription Boxes (a subscription box platform), Amazon Ignite (for the sale of K12 educational materials), having dropped an earlier version of this (Inspire), and Second Chance (for recycling), having also acquired the pharmacy platform Pillpack.

Nonetheless Spark had its distinct objectives: To create social networks in a way that allowed people to discover products.  

Spark allowed it to learn considerably about the way customers create and curate feeds. It also learned the importance of winning brands over to new programs early. By mid-2018, few brands had engaged with Spark, one of the reasons for its demise.

And being a learning environment is very much a characteristic of a Lighthouse project. It also learned about image curation. Amazon’s page descriptions are notoriously difficult to create and Spark allowed it more insight into how people interact with images, how to handle fast-changing image libraries at scale. And how to empower customers on its own platforms.

Live and Explore take the company beyond social networking into something more profound. The world is moving on from images to video. And with Live and Explore, Amazon is attempting to get on board with that change, learning the lessons from Spark (and, in particular, getting brands on board from the outset).

How the Explore platform works

Explore is a live-streaming platform that the average Amazon seller can use. But they have to use it for distinct one-to-one events. For example, a chef in Buenos Aires delivers a cookery guide to a household in Michigan that is preparing for a family get together. This is not so much a lesson, as in many of the online courses that sprang up during the pandemic. It is an actual step by step instruction. In the course of this, or even prior to it, the household receiving the guidance can buy various products (kitchenware, food, wine, tableware) from Amazon.com.

Those doing the “exploration” can also take photographs of the experience, which could also be, for example, exploring Japanese temples, or visiting Argentinian markets, and can speak to locals by activating a speaker/microphone on the host’s device.

The Relevance of Explore as a Lighthouse

Amazon does great ecommerce platforms and engages readers in its powerful reviewing engine. But it is not so good at social networking development, that is getting people to interact with each other through its platforms. 

It’s a common weakness among tech platform owners. But, regardless, Amazon needs to be in that space. It must also take on the challenge of livestreams in ecommerce. Global rival Alibaba, hosts thousands of livestreams a day and the success of TikTok shows that people of all ages want this immediacy.

Having failed to make progress on the first of these (social networking) with Spark, Explore is an attempt to broach the same expertise through livestreaming.

Simultaneously, it has launched Live, which is its streaming platform for brands.

What will Amazon learn from these experiences? 

Livestreaming is a difficult technology because the learning experience is distributed across the host community as well as shoppers, not just Amazon. 

That means Amazon must combine customer experience design with some degree of learning cycles for users at the same time as teaching itself to manage livestreaming in ecommerce.

In addition, whereas previously Amazon’s platforms have been relatively static, or in essence, a vast database of product pages allied to customer accounts, livestreaming is very dynamic, with hosts scheduling livestreams at any time of the day or night, any day of the week. 

In effect, livestreaming flips the Amazon process model on its head, taking it out of its traditional stable and static environment into an “anything goes” mode of operating.

So Amazon needs to learn:

  • How to couch a livestream platform experience in a way that is distinct but easy to use for hosts and shoppers.
  • How to onboard users into a potentially difficult and dynamic experience.
  • How to make the technology completely transparent so that nobody thinks: this is complicated.
  • How to frame the technological constraints of hosting livestreaming.
  • And the constraints of scaling it.
  • What work practices and processes at Amazon will make these host and exploration experiences engaging, including how to manage and adjudicate complaints.
  • How to price services and optimise revenues.

None of these are very obvious. And success or failure will be very public. It is unlikely that Amazon can afford to fail. Ultimately it must become a livestream superstar because that is what the public wants. Explore is the lighthouse for this, the lab and incubator that will help it to grow its skills.

Conclusion

By observing the way that Amazon functions and through our own experience of designing and executing lighthouse projects, we have shaped a model for how to transform successfully. We call it Lighthouse Thinking®

The core of the model are the lighthouse project, the generative operating model,  the learning environment it enables, and scaling down change to allow skills to grow and the future to be defined iteratively. A broader definition would also include the transformation sprint, an agile way of describing a company’s context and the opportunities to overcome dysfunctionality.

A way to shape these opportunities can include other lessons drawn from Amazon. 

The Press Release and  FAQ for a future service that the company uses as a process, means a commitment to thinking through these changes from the perspective of customer success.

But the execution mode is the lighthouse.

Amazon’s greatest successes such as AWS have been built on lighthouse projects. 

This is highly relevant to every company. Amazon should be a model for how to do adaptive business strategy and remodelling all the time. 

After all, we have just described a situation where it is launching a half dozen new platforms in a very short space of time. 

Simultaneously, of course, it is still evolving its hardware platform (Echo) and its AI avatar, Alexa as well as improving its logistics chain and its packaging.

The lesson from Amazon is you can change a lot of things a lot of the time. 

You can develop learning environments that allow you to act fast. There are other dimensions to this that we have not dived deeply into, such as the overall innovation process. But just this one area, the lighthouse project, shows Amazon’s extraordinary willingness and ability to develop new businesses. If you are in a company that aspires to be agile, lighthouse projects are where you need to start.

Amazon has now embarked on a new era of change, again through a lighthouse, Explore. 

Regardless of whether Explore becomes a new megaplatform for the ecommerce giant, livestreaming has to become a core skill and one that is widely distributed inside Amazon, well beyond Twitch/Prime gaming. 

In that sense, Explore is the lighthouse showing the way to highly scaled, user-led activity that leads to new revenues and new shopping experiences. 

Amazon can’t do without it and Explore is the vehicle for the important facets of learning it needs to do if it is to step up its ecommerce game.

Given Amazon’s scale, success and value, these lessons should be fodder for every company and every management team.

Observers of Amazon tend to focus on its acquisitive, predatory nature as it moves into new fields of commerce seemingly at will.

However, it is worth stopping and noting too, that Amazon is also willing to roll its sleeves up and do the necessary learning exercises that keep it relevant. 

Ecommerce Livestreaming is currently in the gift of startups but you have to say, the older incumbent in ecommerce is not being slow to find its own way. Therein lies a lesson for all large enterprises.

Do you want to  avoid the problem associated with big elephant-sized planning? If so, then lighthouse thinking can teach you how to do it differently, at lower risk and more chance of success. If you want to equip your teams to design lighthouses and contribute to your change strategy, then let’s talk.

Browse

Article by channel:

Read more articles tagged: Business Model, Innovation, Operating Model

Strategy & Innovation