Business Models — Sounds like a vague business school term, right?… but when you’re building components, you need to think ahead. It’s a different and less talked about business segment. Today you hear mostly about Consumer Hardware or Software-powered Marketplaces. So I had to learn most of the below from discussion with other “Component” entrepreneurs. Let’s talk through my thought process and strategy to turn our Energy Harvesting component into a platform.
Licensing
I have been doing a lot of research around potential business models for our technology and one of the initial ideas was to license to OEM’s. In discussions with a number of hardware entrepreneurs this seems like a bad idea for creating long term value for our specific situation:
- Prior to reaching product stability, there is generally a large amount of product development required — both on the manufacturing and product design side. In a strictly licensing business model, it would be likely that the OEM tests the product and makes the changes and as a result, harvests the newly created IP value. In addition, our engineers no longer sit at the “core face” of the application — the OEM has direct contact with the end-customers making it difficult for us to determine where to focus our engineering efforts. Over time, this will weaken our position as the value of our engineering team and intellectual property declines. It also makes it difficult to control the path of our technology and can easily lead to commoditization.
- With licensing, we must also be ready for likely litigation. The licensing model is generally not attractive to VC’s and more difficult to exit unless we are acquired by one of our OEM customers. Further, there are a lot of hurdles that OEM’s can put in place for us as a licensee — and we will have expenses related to legal fees/possible litigation for awhile before receiving any significant revenues. There is also risk around a “patent protection” strategy and the ability to protect our invention due to full-disclosure requirements (ie. the licensee will need to manufacture our design). For us, with a single technology, this can become a “huge gamble”. It is a better business model for a university than a single technology Company. Again, we need to explore this further (see the ask around Hardware Strategy in a previous article).
- How do we separate the “Tire Kickers” when licensing? We can burn a lot of time and effort trying to close licensing deals which can take between 6–18 months to put in place. 1 in 10 will actually have the resources to go to market, and then another 1 in 10 of those will actually follow through. This is a tough place for a small Company to be as the longer we negotiate, the less leverage we have. We likely need to have a product that we can start selling immediately in order to maintain leverage.
- I do believe a licensing model can work over time, but in the meantime, we need to also be selling a product, so our engineers remain at the core of the application and have access to end customer feedback. This will allow us to build cash flow to maintain leverage. Once we are selling a product, licensing may be a quicker way to help us scale, but I believe it to be a bad business model initially. I would also like to further explore the ARM Holdings model as they license 100% of their designs, so the question is how did they deal with the challenges listed above and could we copy that model (would love to hear relevant feedback here)?
- ASK: If anybody has a contact with deep licensing experience in the “Hardware Component” space — we would like to start discussions around the points above.
Semi-Custom Business Model — ASIC’s in the 80's
Based on the above it is clear we need to “Sell something” to continuously build our IP and gain insight about the market. So, how do we turn our product into a “platform” — a component that can be customized and built upon in products across multiple verticals? This is more cost intensive and contains more risk in the near term than a licensing strategy, but can provide a much more valuable and defendable position. I also believe the cost/risk factors can be managed. One method to follow is outlined below — I saw this model used by a lot of the component technologies at the Conference, and also on the Witricity, Freevolt and Energous websites.
3 Prong Strategy:
- Stage 1 — Development Kit: One possibility would be to sell a basic “Energy Harvesting” development kit including an antenna triplet and AC-DC rectifier to convert the ambient RF energy into DC power. We can sell this in a specific size that Companies can use to hook up to their sensors, IoT devices, or electronics to demo a basic version of our technology. A development kit would bring in a small amount of revenue and reduce the amount of “Tire Kickers”. This is a low-friction way to bring customers into our sales funnel.
- Stage 2 — Consulting: Based on customer specs (power requirements, form factor, cost, etc.), we could work with OEM’s to build customized solutions using our initial development kit and then charge consulting fees for these customizations. This strategy works well as OEM’s fund the development of our IP and provide access to their end customers while we are still a small Company. This is a great way to scale our development and engineering expertise with a lower cost structure.
- Stage 3-Launch Mass Market Product: The end goal of Stage 3 is integration into mass market products of the OEM — as a result of the consulting and development we have done. There are a number of items with this strategy that make sense: Initially, we can avoid a lot of the issues related to dealing with a large OEM because we are selling a “lower friction” development kit at a low price-point (ie. due diligence on our manufacturing partners and support, due diligence on our current customer base and MGMT team, complex contract structure and licensing negotiations). We can also leverage any development across our entire customer base and have revenues right from the beginning. Because we are selling a brand new technology, there will 100% be problems. This gives us time to develop our manufacturing “know-how” and time to develop solutions in a low risk, low-burn environment. As opposed to the risks in the licensing model detailed above or in the full blown “Product Development” model building a commercial product from scratch.
Other Items to Consider
- Selling a demo kit to Tier 1 players and larger OEM’s can be difficult. Some just won’t buy. To jumpstart sales of “Development Kits”, we need to first put together a demo that we can show and then execute on classic business development strategies (ie. conference demos, free demo’s for targeted influencers, etc.). Based on initial reactions, we can reevaluate the above strategies and need for “Dev Kit”.
- A way to drive sales of development kits is to tell them we are already selling to their competitors. For many of these Companies, the speed to market is the only advantage they have, and the biggest “pain point” is adding a new feature to their devices before their competitors. We need to portray the image that we will be “Faster than competitors and their internal development”
- Dealing with a Tier 1 at this stage will likely be close to impossible (ie. Apple / Samsung). For example, Apple is very tough to engage, notoriously secret and often restructures — so it is quite possible that the person spearheading our project is moved out while we are discussing the project. Even if we get to the point of discussing a deal, we have no manufacturing chain and minimal test results. This is not a good early customer for us unless they are an acquirer. They can also be extremely dangerous if they decide to tie us up with a lawsuit, back a competitor or replicate our technology. One path would be to try and scare them by doing a deal with Tier 3’s first and then to approach them.
- ASK: We are looking for Entrepreneurs that have sold or licensed a technology to Apple or Samsung to talk through best practices and what would be required before approaching.
- We are better off targeting Tier 2–3 partners that are more desperate for new technology and more flexible to work with a smaller provider. Targeting these type of customers first will give us time to sort through our development issues and tighten our production know-how. Ideal customers would be between $100m — $1Billion market cap. We should approach these customers first.
- The Component Strategy mentioned above is a difficult grind to start with. We will need to fight for every scrap of business initially and it can take years to crack through. OEM’s generally don’t want to take a risk on new technology. We need to build manufacturing and a customer base and this is very difficult as we test and iterate on our technology. For example, the sales cycle described above can easily take 6–18 months to go through (Step 1–3 above). Sometimes a single PO can be negotiated for 6 months — and that is for a complete and stable product. I imagine our initial products will fail a lot of mandatory spec requirements from OEM’s. However, once this cycle starts it is virtuous and generally accelerates quite quickly, with more customers, more revenue and IP development, more demand, etc. Word quickly gets out among OEM’s and risk decreases as you work with more stable customers.
- Another thing that needs to be considered with a demo kit is the risk of technology theft/replication. We have no software/brand/community to protect our hardware, so we definitely need a few more opinion’s before sending anything to customers. I am working on this right now with engineering/manufacturing experts in Israel. (Please see initial ask on protecting technology).
Alternative Models: Building A Product From Scratch
There are a few other Business Models to consider — if we had a niche market identified with a clear use case that we could deliver on, I would lean towards this despite the risk. However, I think we are still trying to “develop” our product towards this stage utilizing financing from potential customers as per above model. This would also be a suitable model if financing was plentiful!
- Manufacturing Partner: Partnering with a large CM to help with product development and provide legitimacy around capabilities. Potentially negotiating investment from the CM in return for reduced “NRE” and “Tooling” fees. We may partner with a Jabil or Flextronics here for help with antenna design + electronics miniaturization. This strategy also carries risk (IP + Flexibility, etc.). We should continue research here after speaking with initial target customers.
- Developing our own Branded Platform / Killer App: Some other strategies discussed included a “stand-alone, plug and play” energy harvesting kit that can be added onto any sensor based on energy, size, cost and environmental requirements. Effectively we would build a platform that could replace the battery. This business has enormous risk — You have all the risks of hardware (ie. distribution, manufacturing, inventory management, logistics, branding) and to some degree you can become dependant on a small group of providers of the underlying device (ie. sensors), that could potentially block your device, change their form factor, etc. The multiples on accessory businesses are extremely low and these are often “knocked off” very quickly. Accessories generally only work if they are an “upsell” to “your” underlying device. However, in this case, developing a platform that can be used across all wireless sensor networks reduces the power of the sensor manufacturer. We would also be selling direct to the sensor manufacturers as an added value solution instead of circumventing them and selling to the end customer (ie. like in the example of the Mophie). Further, becoming the “defacto” platform for power can result in mass adoption and customer stickiness and huge multiples.
- Drayson Technologies has recently developed an interesting product: They have a self powered Carbon Monoxide sensor that acts as a proxy for air pollution. It is a credit card shaped device that connects to your smartphone via bluetooth. The device provides live pollution and air quality readings to your smartphone which are updated to the cloud for others to view through the app (ie. very similar to how Waze uses crowdsourced data for traffic patterns). Air Quality is now one of the leading causes of death in the world (ie. worse than AIDs + Malaria combined). The App details air quality around London and is gamified to give users rewards for travelling on “non-polluting” routes via bicycle, walking or bus. It is easy to imagine this live pollution data being licensed and monetized. While I think this is an interesting business model, the hardware piece costs $70 pounds and is a bit bulky to carry around all the time, so the pricing model may prevent adoption initially. I could also see these sensors being included in all mobile devices in a few years along with plenty of structural IOT devices — so it is unclear if they will be the first to scale and create the largest “live-pollution” monitoring network or if other devices will build the network first. This example was more used to show that another option for us would be to create a product focused on the end user with a community, software and valuable data to build a protective moat etc. as opposed to a “component” only strategy.
Overall — we will need to consider all of the above while prioritizing and developing our product features. I’ll get back to you after additional research and product development.
