In 2014 the Bluetooth beacon ecosystem has evolved and grown substantially. It promises explosive growth over the next five years. We can now apply some structure to understand the mass of new companies that are competing to win a race to bridge the digital to physical divide. In this article we outline a draft of what we are calling the Beacosystem Technology Stack, designed to help make sense of a very important part of the Internet of Things.
Hype is an explosive but necessary ingredient in the rocket systems required for any new technology to achieve escape velocity. The gravitational forces of fear, uncertainty and doubt are powerful. Without a massive injection of hype, entrepreneurs, VCs and early adopters will fall to their doom before the pilot solutions can prove that the way is safe for the majority of industry players to join them. The beacon ecosystem is particularly challenging because while it looks like high tech, it requires brick and mortar deployments that move a lot slower than software deployments.
In 2014 the Bluetooth beacon ecosystem was pulled into the back of its bucket seat with the thrill of major G-forces, sufficient to put indoor location into a stable orbit. As a result we have seen something of the order of 100 beacon startups get funded, racing to find their niche. Giants of the industry Apple, Google, Qualcomm and Samsung have placed big bets and launched key offerings. The early pilot projects by the likes of Major League Baseball and Macys, are now transitioning from pilot to full deployment.
For any market to be funded, business plans are required. They range from boring (not funded), exciting (funded), crazy (not funded) and insane (acquired by Facebook). To develop these plans, business development professionals engage in a ritual of arm wrestling with their finance counterparts. Given the complete lack of certainty that most of these plans are based on, these contests should result in almost every proposal being canned. The only thing one can say for sure about any of these business plans is that the numbers will be wrong. It’s just not possible to predict early stage business results with certainty, while weathering a storm of unknown variables and unexpected events. Fortunately, we have a solution to this problem; the industry analyst. In the beacon market, one of the best analysts, if not the best is ABI.
In June of 2014 ABI predicted that total deployments of BLE beacons were forecast to grow from 65,146 in 2013 to over 60 million in 2019. The only numbers we can be sure of in their prediction are “2013” and “2019”.
In other words, its highly unlikely ABI will have got it exactly right, but the numbers are big enough to be interesting and ABI will have done a better job than pretty much any one else in using a credible process to produce the forecast. The analyst acts as the whipping boy of the boardroom. In the same way that royal children in days of old had a whipping boy whose job it was to be soundly beaten when their master got things wrong, so the industry analyst is prepared to brave the probing questions of dubious investors and produce numbers that would be the source of anger, hilarity and disbelief if produced by someone that worked in the same building as the people reviewing the presentation. So the BD guy and the finance guy agree to suspend disbelief, and the plan gets blessed.
When these 100 or so beacon startups sprinted off from the starting line, their first products looked remarkably similar. An awful lot of beacons were running off coin-sized CR2032 batteries. Most of the battery life predictions were even more optimistic than the business plans referenced earlier. So beacons were “conking out” prematurely. The payback for these “performance problems” was that sometimes this would occur during demonstrations to prospective customers (“… this has never happened to me before”) or after a lot of work had been invested placing the beacons in hard to reach locations … across a whole bunch of stores in far flung places (“you do realize we just spent the last month putting these up”). The poor engineers who had been forced to produce the battery life projections weren’t to know.
Houston, we have a problem
The battery life of a beacon is impacted by multiple variables, which were changed as we figured out what was needed in the real world, including:
- Strength of signal being broadcast, which loosely correlates to the range it can reach.
- Frequency of transmission, faster speeds – for example broadcasting 10 times a second consumes more power, but is required if you want your app to see the beacon when it’s in background mode or there is significant movement between the phone and the beacon.
- Packet size – iBeacon packets should theoretically all be the same size across vendors. If you want to leverage all the important features that Apple puts into iOS, and if you want to use the iBeacon logo on your product, you have to transmit a packet that conforms to the iBeacon specification. This is relatively simple, just three numbers that establish the identity of the beacon and a signal strength indicator to let recipients know if you have turned the signal “volume” up to 11 or not.
Unfortunately enabling anyone to use your beacons in this open and unprotected way has proven to be inadvisable for many applications. If the ID numbers are simple static values, anyone can clone them. If you were using the beacons to authenticate a customer’s presence in the store before awarding them an incentive, a devious hacker could replicate that signature and collect the rewards from the comfort of their dorm room. Also, many retailers don’t want mobile applications from their competitors to be able to identify the fact that their customer is in a specific location in their store. Armed with that information, the competitor could then persuade that customer to buy the product in front of them for less on-line.
So how do you retain your Apple iBeacon compatibility while obfuscating the true identity of the beacon? What many beacon vendors have done is to set the iBeacon IDs of all their beacons to be identical, enabling their app to use the hooks in iOS to wake up when they see a beacon but not betray the unique location. They then transmit an additional encrypted packet that contains the true identity of the beacon in a way that only authorized applications can understand. Unfortunately, when you double the payload of information you are transmitting, your battery life drops accordingly.
So very quickly beacon providers have had to bulk up their battery packs, field update their beacon firmware and introduce a packet structure that is as standard as a snowflake i.e. not standard at all. Such is the charade of high tech standards. Anyone remember the SQL standard or OSI?
Figuring all this out, dealing with hardware supply chain issues, customer support, developing a management system that makes tracking and adjusting beacons in the field possible, dealing with multiple handset OSes that change behavior on a lunar cycle, deriving meaningful analytics, attracting developers, selling to brick and mortar retailer customers (with a core competence in the art of procurement torture) with limited time and patience … makes the beacon business a lot harder to win at than most people thought.
The result is that those beacon providers that haven’t achieved the volumes to establish significant market share will have to find other areas to make their mark. The survivors will have to focus on vertical markets (venues, home automation, hospitality, personal item tracking) or specific parts of the stack.
The Beacosystem Technology Stack
If we can agree what some of the components are, it should allow us to better evaluate the capabilities of the players and figure out candidates for best of breed and understand the direction those players may take. The categories we have considered here do not include every component that a solution containing beacons might have. That would require us to describe everything. Most beacon systems will incorporate databases, but it doesn’t make sense to detail those there. These categories are intended to describe the places where new entrants will have to create something beacon specific to service the proximity market place.
The Beacosystem Stack is interesting because while the components are specific to proximity based solutions using beacons, the same components will likely be used with other proximity technology. iBeacon is acting as a catalyst to incubate an ecosystem that other proximity technologies such as audio, RFID, LED/Visual Light Communication (VLC), magnetic resonance and LTE can benefit from. These technologies may have pre-dated iBeacon and not achieved the level of momentum to assemble the equivalent ecosystem, or like VLC they may still be embryonic. In this way iBeacon has spawned and invigorated a broad category of technologies and providers that may outlast what we know as the BLE beacons of today.
It would take a book to comprehensively review each of these Beacosystem Stack components (we are working on that).
So instead let’s look at a few players and match them to the framework.
Based in Poland, with a brilliant flair for industrial design and an ability to market themselves without actually having many (or any) marketers on staff. They started off entirely focused as a Beacon OEM, by doing that they became the defacto leader as they didn’t compete with others building out other layers of the stack. Given that the only folks making much money on beacon hardware are the Chipset vendors, it has made sense for them to build out a Fleet Management capability.
Have offered multiple beacon models from early on. An entry level S10, which runs off a coin cell battery (the ubiquitous 2032 that has been widely adopted by beacon manufacturers). They have used this to engage developers and for applications where small size and low cost is important Vs longer battery life. The S10 is the lowest priced beacon on the market at $5 for single quantities. It was originally designed to fit on shoe laces and in back packs for the tracking of children as part of an amber alert project. Their larger beacon was designed for applications where battery life needed to be maximized. The S20/21 runs off of four double AA batteries and is the model that you can still see in use at Apple’s retail stores by their own shopping app (look carefully behind the “J” hooks and under the accessory shelves).
- Radius Networks
Have been focused on the hardware and API components of beacon systems, with a lead in USB powered beacons, making a play in the standards area with their Altbeacon standard which offers APIs and a specification for beacon packets that could work across Android and iOS. As a consequence exercising good “lane discipline” (not sticking their fingers in other people’s lines of business) they have been a go-to-partner for the likes of Samsung, who are aspiring to dominate a large portion of the rest of the stack.
Lighthouse and Urban Airship – are examples of Campaign Management providers unencumbered by producing beacon hardware. Lighthouse built their offering from scratch, while Urban Airship extended from an adjacent market (SMS campaigns).
Have been working in the proximity area for years with their gamified shopping app. As a result they have enjoyed a significant lead in signing up and deploying beacons as a fast-follow upgrade to their sonic hardware that authenticates shoppers’ presence in stores. With millions of users already using their app prior to the existence of Bluetooth beacons, ShopKick probably has more production users than any other player now that they have switched to that technology. Since the acquisition or “merger” with SK Planet, they also have a huge advantage being one of the few Beacosystem companies that merchants don’t have to worry about Google buying. The last thing merchants want is to find Google beacons stuck to their shelves. The question is, what will ShopKick do with that beacon infrastructure sitting in some of the world’s prime retailers. Will they rely on their app to monetize it, or open up access to third party apps in partnership with their retailers?
Now let’s look at this stack from the component view
Campaign Management – Arguably you could put content management systems (CMS) in the stack, but we believe that most beacon systems will leverage and integrate with existing lower level CMS components that can work across channels.
The CMS is typically leveraged in a Campaign Management layer that orchestrates actions to deliver compelling user experience. Examples of vendors in this space are Rover Labs and PassJoy. PassJoy addresses the challenge of how to leverage beacons without needing an app to be installed on the handset (traditionally having an app is a requirement). They do this by orchestrating a link between a targeted Facebook Ad that then provisions a pass into Apple’s PassBook. The pass can drive alerts with simple marketing content and offers to the consumer as they approach a beacon. Ultimately the passes can upsell the consumer on downloading an app with much deeper functionality. Rover Labs have started off with a few simple APIs to allow developers to build those proximity apps, abstracting out a lot of the complexity of the lower level beacon APIs and managing the content to fuel the apps.
Beacon Networks – encompass registries such as Radius Networks’ WikiBeacon, where you can see the location and number of beacons installed around the world by country, city and street. It’s just a sample of the beacons the Radius app sniffs out, but an interesting indicator of where the action is. The Beacon Networks category includes vertical networks from Sita, a major provider of solutions to airports. Sita provides a registry, standards and APIs to service the needs of Airports, focusing on use cases specific to that environment, check-in, navigation to gates and finding bags. Gimbal have established their own public and private network functions within their product, using their proprietary/secure packet technology to allow developers to view maps of cities and regions and see public beacons they might leverage in their applications.
Payments – PayPal was an early mover in announcing their beacon and positioning it to primarily be used with their own app, rather than as a general-purpose offering. Unfortunately, they have yet to publically announce pricing and availability. CurrentC, the merchant consortium formally know as MCX, is developing a cloud based payment system that would work well with beacons as an alternative to QR codes. They seem to be focused on getting the basics done with QR but could offer their members and consumers significant benefits if they allowed them to pay via beacon, without the need to scan a bar code. When Gimbal piloted this beacon triggered payment app at a local coffee bar, the user experience was phenomenal, gaining a huge share of the transactions within just a few weeks.
Ad Networks are a particularly interesting layer in the stack, one with probably some of the biggest revenue potential, but trailing in maturity. This is understandable; the advertising market can only really take off when you have beacons widely deployed across the country. Advertisers are not going to be interested in buying access to a network that encompasses one or two DMAs (Designated Market Area). Launching a beacon ad network is like throwing a party, you need lots of different people to show up at more or less the same time or it will be a dud. Very few people want to be the first ones to the party, making awkward conversation with a host who is hiding the panic in his eyes.
There are quite a few contenders for the ad network space:
The OS providers
Google – This is an extension of Google’s core business. They are really well placed, in terms of critical mass, knowledge, customer base, assets such as the Androids OS and the Chrome browser. One of their biggest challenges is that the retailers hate them. That’s a strong word, but retailers are really suspicious about Google acquiring their data and selling ads that drive them out of their stores. It totally makes sense for Google to buy some of the other smaller companies who can innovate in this space using their platform rather than build something from scratch and do a frontal assault.
Apple – Like Google, Apple already has so much of what it needs. With Apple Maps it could offer very relevant ads as to wander around any shopping center in an app which we all have installed on our iPhones. iAD allows them to surface location specific ads in apps. You are playing angry birds while waiting for your flight and you are reminded of a Ben and Jerry’s, 30 feet away from your seat. Passbook is a preinstalled app that you can’t un-install. It is already surfacing location-based promotions using GPS and iBeacons. If connected it to an ad network it could do a lot more.
Uber apps (major apps that have a ubiquitous presence on phones)
Facebook – Check-ins get more precise and once there is a critical mass of beacons, expect those Facebook ads can get even more relevant.
Specialist apps – Companies like GasBuddy have millions of users and a clearly-defined segment that they dominate (people looking for cheap gas). They are well placed to monetize the real-estate in their apps working with brands with which they already have relationships. With promotions that are trigged as drivers approach the gas pump, GasBuddy can convince more people to go from the forecourt into the C-store, to buy the snacks that fatten the wallets of the station owner and the waste lines of the driver.
Existing ad networks
Outdoor advertising companies like Titan and JC Decaux already have the physical presence (billboards, digital displays, phone booths, bus shelters), the sales force, the base of customers and the knowledge. It makes sense for them to add beacons so that they can sync the ads on your phone with the billboard we pass by, tailor the products presented on the digital display and provide better attribution and segmentation to their traditional advertisers. Even if a targeted ad doesn’t show up on your phone, or on a billboard, it’s valuable for them to understand the kinds of people who walk by their displays and who stopped to view them.
The challenge that most of the contenders for this space have is that they are big. This means that they are inherently less flexible and struggle with innovation in the face of politics, along with limitations imposed by their existing partners and customers. The number of bets they can place is limited. IBM was once famous for setting up development teams that competed with each other, as a way to mitigate this issue. That’s something we don’t hear about so much anymore. It makes sense for the big guys to partner with smaller companies and then innovate through acquisition.
inMarket and Swirl have both been active building Ad Networks. They have developed comprehensive offerings, delivering their own hardware, analytics, fleet management, campaign management systems, and turnkey apps to retailers in order to get their beacons in store. They then partner with app developers with real volume to get eyeballs. While both firms have done a nice job, their model has some scalability challenges; especially given the breadth of components in the Beacosystem stack they are providing themselves (most of the components it seems).
Unacast are one of the newest entrants to this space. Their approach is interesting because they have focused quite narrowly on privacy and a hub that works across any number of Beacon OEMs, Beacon Networks, Campaign Management Systems and traditional ad-networks in order to retarget online campaigns using the insights gleaned from the off-line brick and mortar behavior of the consumer. Their model adds functionality and shares incremental revenue to players outside of their area in the stack.
Other companies have been focusing on retargeting the customer in the actual proximity app. Unacast addresses the opportunity beyond the mobile app. The retargeting they do is across all digital properties (web sites, displays and other applications) where the end user sees ads. Successful engagement requires repetition and consistency to be effective, so this makes sense for the advertisers in terms of driving conversion rates. For Unacast and their partners it opens up a much bigger market than click-throughs from a single app.
In this last example, orchestrating a campaign across multiple channels takes an ecosystem to achieve. Any strategy where you can make money and enable an ecosystem of partners to make money too has major potential. Given the velocity of this market, playing well with others is essential. You can’t build and market everything yourself, the risks are too high and there isn’t the time. The make versus partner decisions in the Beacosystem Stack is critical. Some companies are clearly inclined to provide more of the pieces than others. Apple, the company known for providing complete solutions, has only provided a few of the many components required. In doing that they have inspired an army of others to fill in the gaps in the framework that we have outlined above.
Getting into orbit is just the first stage in the journey. Just as hard is landing safely. In a Beacosystem Stack that is large, complex and competitive, those companies that figures out where to focus and how to partner well with companies in adjacent areas will be the ones most likely to make it home safely.