Intangible.Tech

This article has been a long time coming. Through a whirlwind of experiences over the past few years I’ve been exposed firsthand to how large organizations work and operate. Operate being the primary focus of this piece and the most critical piece in how organizations are going to shape themselves moving forward. Branding, technology, product, funding and market value are all addressed with the main point being – you can’t simply look at just one area to realize the value of a company or its products.

At this moment Tesla’s value is more than double that of major automakers such as General Motors and Ford. To outsiders this comes as a surprise. Looking towards my hometown of Muncie, Indiana – a town built and then quickly reduced to near nothing as the auto industry has shifted – this has been something chipping away at my mind for a long time and why I’m using Tesla as a centerpiece to highlight what’s happening or going to happen to every company in this new age.

The Power of a Brand

Branding is everything. Working alongside the world’s largest and most famous brand creator – Procter & Gamble – I’ve learned to appreciate this and respect it more than ever. As someone who has always heavily researched and bought products (excluding clothing, a topic I’ll address later as this is also shifting) based on specs I never cared about branding as much as what the product can do for me. However, I’m an oddity. Any flaws in design and build I could fix myself (the early Android days were a shining example) but to the normal person they expect everything to work and be intuitive to use. This is exactly why UX and design are becoming so critical to digital experiences – areas that have always been historically covered by their physical (tangible) counterparts and are just now gaining the popularity they deserve – see JP Morgan’s news about increasing their UX department 3x since 2016.

How does everything work, though? For many brands it’s through small feature additions that can be marketed. There’s little value in advancing something when a brand can simply market and give the perception of being a better product – if something is good enough to be marketed, that’s what they’ll do. Consumers then familiarize themselves with the value the brand conveys through marketing and placement – not the smaller details like company B’s (with a fraction of marketing dollars) better solution to problems (product). This has worked very well in the days of little information available towards a product.

However, as digital has now risen to the forefront this model is at inherent risk. It’s not because these brands have lost their marketing punch or consumer recognition – it’s because their products and the way they work are fundamentally changing as digital requires new ways of working, thinking and business models that support them. In order to have digital experiences that consumers have come to expect, everything needs to be connected and digitized appropriately with the focus towards the overall consumer experience and not just towards projected (guessed) experience.

Minimal Viable Product & Scaling Strategically

Looking back towards Tesla, the entire evolution of their company has been rooted in breakthrough technologies. In the early 2000’s no company had ever been able to produce batteries that would allow an all-electric car to be feasible. They started there, developing batteries and drivetrain technology that had to be created internally. Because of their constrained resources they looked elsewhere for the additional pieces – Lotus in this case whom provided the chasis and other traditional auto components that Tesla at the time had no resources to create.

In product we would call this the MVP or minimal viable product approach. It allowed them to get something into the marketplace to prove viability, marketability and most importantly highlighted to investors that their continued investment would yield continued improvements.

This is where I believe Tesla made its ultimate strategic decision that’s allowed it to reach where it is today. They could have simply gone the route of licensing their technology to other brands – in the case of the Roadster Lotus could have sold the electrified car under their brand making Tesla the supplier. Instead, they opted to show the world they could be in the driver’s seat towards a better car – a longer term play now paying dividends.

Modern Product Thinking – Approach the Problems from a New Angle

To do this they needed to re-think every component of an automobile. Battery technology can only move so fast so what else can be done to narrow the gap between combustible cars? Retrofitting the electric chasis onto the Lotus frame lost efficiencies. They needed to create a purely digital automobile where all variables and elements of the product can be monitored, controlled and improved while still being mechanically sound.

In businesses today Tesla’s scenario is common. For manufacturers they look for supplier that can build a part that will fit into socket A and then another supplier who makes socket B to complete a part or segment. This is purely mechanical. For IT departments or tech companies you look for suppliers (vendors) who can connect their product to the rest of your components digitally. Variable A connects to Variable B which is then one of many (digital) cogs in a very complex system especially when scaling is introduced.

Supply chain networks while somewhat digitized are still prone to shifting variables and unforeseen delays. A single missing part could delay an entire product months despite everything else being in place and working. No matter how much digital work you put into this – if any part requires analog (real human) interaction there’s going to be bottle necks and the amount of retrofitting technology into gaps is costly and mostly a wasted effort.

Typical Vendor and Supply Chain Processes Across Large Enterprises

IT departments and modern tech companies grind to a halt when data sources are altered or services shut down. In legacy organizations this could mean (gasp) mainframes or even COBOL (ancient) code dependencies being relied on to work effectively. Modern architectures and automation have largely made many of these issues extinct but unless your company has kept up (they haven’t) you will still be at the mercy of the same supply chain constraints of the physical world.

This is where Tesla approached their issues head on through the practice that is best summarized as modern product thinking. Tying in lean development, design thinking and agile development practices that when combined, solve problems (product) in the most effective way possible. Although legacy solutions to problems may be pitched, it’s ultimately an empowered team from various backgrounds that comes together and drives innovation forward.

You Can’t Innovate with Suppliers – Where Innovation Really Happens

How does this contrast with the way sourcing has always worked? Well, traditionally companies will outline specs and go to their suppliers with something that needs built. The supplier will take that back to their various teams and have them pitch the product to the client (company) and that they can do it for X amount of money – typically a wild guess based on a purely financial analysis, not manufacturing capability. The company compares the work against other suppliers who may be vying for the contract and then the legal agreement is then made and (hopefully) by X amount of time the company receives the product on an ongoing basis with the supplier.

All the supplier needs to do is build to spec and in a year everything will be as it should. The primary brand will place project managers with the supplier to ensure they’re up to pace with QA and other engineers ready to jump in and tell them how they’re not doing their job should they falter. Sounds great and easy, right?

Wrong. With this process anytime something changes (specs or demands) the supplier will need a new statement of work, legal and sourcing sign off just to continue the change costing time and money. Imagine the legal, engineering and accounting hours you waste 1) going through this once and 2) going through this after you inevitably make a mistake.

Business is a game of risk mitigation. Legacy companies have approached reducing their risk profile for the above scenario happening by having extensive testing through what is known as Waterfall development. Research and development is done internally having to pass through many gates before ultimately reaching the point of being handed off to a supplier. This is done to help ensure nothing changes as manufacturing is set to begin which in the case of brand-new technologies and innovation – could take years if not decades. The cost is then that you spend X amount of time and money on teams getting through internal processes while nothing is getting released effectively providing zero value to you or your customers.

You may, however, be putting out PR to artificially boost your brand’s perception of innovation catering to your shareholders – a tactic proven to be insignificant and any jump in share price to be negligible if best. If anything this is a sure sign of struggling internal process.

Waterfall works alright when you have a product that you know is going to sell and where innovation is expected to happen slowly. Jet engines and their sometimes 40-year life spans and even models of cars where their style will remain the same for 4-8 years before changing are examples of this.

In the examples above companies traditionally have business models around those innovation cycles and in the examples of cars, a first year model may have issues with smaller parts that get fixed in later models and consumer delights such as new colors or types of seating being offered. In the case of the small parts needing fixed, brands protect themselves with short term warranties to offset these errors to brand perception. However, these warranties although weighted based on data are not coming from live data but instead, projections.

More critically consumers unhappy with the baseline features must wait for the fix to go into place (short term) or if unhappy with a major component wait until the next model year (long term) or next waterfall cycle. Either way, a loss for consumers and a primary risk of losing brand loyalty.

How is Tesla different?

They shifted as much development as they could in-house across both the physical and digital world. Instead of approaching sourcing as an area to reduce cost by outsourcing development – they viewed it as an opportunity to reduce cost by reducing the overhead of managing suppliers and enabling consistent innovation needed for them to move fast, get to market and delight consumers on a rolling basis.

Again, they absorbed the cost of internal development teams where most companies opt to outsource and put all risk in the form of a contract. This, again, works well when given work is known, well documented and well established. For any new innovation, though, you need people bought into your idea and company, incentivized appropriately and given the autonomy to work as best they can and not be limited by process. This difference in thinking is best summarized by product extraordinaire, Marty Cagan.

Look no further than the various features Tesla puts out within weeks of hearing consumer feedback to see this in action. A dog mode for keeping your pet cool in the summertime, sentry mode for recording vehicle break-ins and ongoing range improvements are just a few of the features that were added to their vehicles after release. In some cases, a Twitter comment to Elon Musk was turned into a feature within a week to users.

How in the world can any automaker say that they gave consumers a brand-new feature in a week’s timespan? They can’t and they never will given their continued operating model which would entail a new statement of work followed by countless sourcing exercises and then, finally, a contractor doing the work which assumes they know how the system works and operates. Incredibly inflated.

Look Towards SaaS as the Golden Standard

The above scenario of getting to market quickly and adding consistent, rolling feature releases isn’t new – it’s common place across the Software-as-a-Service landscape and has a term known as Product Led Growth. Most notably you can look to the gaming industry posting record profits with their products. They fundamentally flipped their business models where the initial product is free to users but for those same users to take advantage of added gameplay they’re needed to pay as they go. To many the initial release of these games are met with lackluster fanfare and riddled with bugs and enhancements.

However, these companies’ primary employees are their developers. They’re paying for them as an ongoing cost – not just a statement of work to release, stop and start again after internal assessments are created. They keep building and iterating – always.

For Product Managers at these companies they will now have data from all areas of the product identified for improvement. But, due to a finite amount of dev power for any given time they must continually make tradeoffs on what to develop next and what is going to provide the most value. Do they fix the bugs effecting 5% of people or do they build a new feature that may grow the user base 400%? Ultimately this decision lies with them and business stakeholders but they beauty of this approach is that they’re continually releasing and making what may have been seen as an imperfect product at launch turn into a product consumers can’t stop using in record time by consistently trying new things on a rolling basis and always integrating feedback in a timely manner.

In a time where 65% of buyers rely more on peer recommendations than pure marketing, it’s vital to continually show to users that your product is not only better than competitors but that it will continue to get better as time goes on. It’s now becoming expected.

Tesla isn’t SaaS, Though – Enter the World of IoT

The Internet of Things is a word many love to throw around as the technology to change the world. Well, it already is changing the world and the reality is it’s a lot of different technologies that bring this to life. In the case of Tesla they’ve made the entire car an IoT device and on a more micro scale, they’ve made nearly every component an IoT device from the ground up.

In the IoT world both physical parts (hardware) and software are blended. Software controls what the hardware can do and likewise when hardware variables change, the part can tell the software to do something different. An example of this is Tesla’s use of sentry mode using cameras located around the vehicle. When movement is detected the camera detecting motion captures a video, saves it and then passes it to Tesla’s cloud service for a user to view anywhere and at any time.

For this to work effectively hardware needs to be aligned at the earliest stages of production with the agreement of software teams. The hardware needs to be future oriented enough to allow software developers the freedom to continue tackling (unknown at the time) consumer demanded features and upgrades. Tesla has this ability built in-house and has the entire car’s components connected right from the factory.

This contrasts with legacy automakers using many suppliers across different vehicle lines. One supplier may provide multiple companies with components but each company requires different software standards and although possible to connect, increases complexity and likewise cost to manage and implement effectively.

An example of this may be Cadillac which has cameras installed around the vehicle for autonomous driving. The same feature Tesla advertises. However, the self-driving software and hardware isn’t GM’s – it’s a GM supplier.

The consequences of this are that say GM wants to create a sentry feature to parity match Tesla:

They now have to engage their camera supplier (whom specializes in self-driving, not cloud recording)
Allow the cameras to be used for something other than they were intended
Detect motion (to go a step further I’m pretty sure Tesla added machine learning to identify faces)
Record back to the car’s main computer for storage
More than likely that same computer used within the Cadillac is using yet another supplier for creation which means it now needs to support the feature of recording motion and receiving the video file
Saving it
Connecting to a cloud service for upload which also may be another IT vendor
Hoping all the separate vendors system’s work in unison with each other

All said and done this would require technical leads at GM to spec out all pieces for suppliers. Engaging all suppliers for said work, opening a new contractual agreement for development and then a project manager and QA manager capable of ensuring all pieces are working correctly together. All in all this process would take months if not a year to implement correctly.

The Red and Green lines Showcase the Amount of Pre-Work Required for the Different Methodologies.

The other caveat? Those individual pieces would more than likely require firmware/software updates and because they’re, again, from different suppliers this more than likely would never come in the form of an OTA but instead in the form of a new model year of vehicle due to the various complexities.

Tesla on the other hand has these core software components baked into their native, in-house software developers. Even if separate teams are needed to fully develop the feature, the teams can work internally, engineer to engineer to solve the problem vs a project manager trying to tie many separate pieces together. Teams are given the opportunity to fail and move quickly vs attempting to over-manage and aim for perfection.

Highlights How Waterfall is so Slow You May End Up Missing the Mark on What Your Consumers *Actually* Want

How Does This Impact Innovation? Suppliers Still Need to Manufacture

In the scenario above there still exists an opportunity for something to malfunction. Let’s say roughly 5% of Model 3 owners are experiencing a flat tire around 12,000 miles all on the back left tire. Because every component is digitized from the start an analyst can pinpoint the issue within minutes (ok, maybe a few hours) and determine what was going on. It may very well be a hardware issue such as a bad strut mount on a part from ‘supplier ABC’ during a particular supply batch. Tesla can now proactively alert customers (showcasing enhanced customer experience) but they can also adjust other variables via software that may alleviate the issue remotely.

A cost benefit analysis would have to occur to determine the most cost-effective solution vs just replacing the part and spending money on a software fix but Tesla unlike conventional brands can adjust other components to offset the part from other breaking. Or at least until the customer can make it in for a repair – by reducing something like the back-left drivetrain to deliver X% less power reducing strain on the device.

Tesla’s Operating Structure is the Holy Grail of Modern Business Operations

This is a small-scale example, but Tesla’s real innovation power is that on a dime, Elon Musk and team could assign a dedicated team of problem solvers to any consumer facing issue and they can have a result in the span of days or weeks. Thanks to software and the connectedness of every piece of the car this fix can be sent to consumers, instantly solving critical fixes or adding value to consumers – that’s the true power Tesla has over legacy auto manufacturers. Bureaucracy is cut to a minimum thanks to processes that only the latest technology and management styles can create – true agility in the marketplace that only a digitally native company can provide.

‍

It’s worth noting that if given the chance for me to purchase a Tesla at a given price point I more than likely would opt for another vehicle such as an Audi S5 or equivalent. I am in no way a die-hard Tesla fanboy despite admiring how they conduct business.

Although slower and non-electric I do believe having a full cockpit and digital clusters as more premium than Tesla’s single screen solution. If Tesla adds these it will increase complexity of their software stack. Their analysis would need to be on deciding whether the added complexity and likewise slower pace of change would be worth the potential growth for consumers like myself.

I also happen to love the sound, smell and ability to customize petrol engines. This may be one area where although I fully admit the superiority of electric, I’m finding myself uncharacteristically old school.

‍