RFP: Sportstech idea dump
Request for product piece exploring a handful of B2C and B2B sportstech ideas with a focus on betting...
Every two weeks I write a piece on sports betting, gaming or trending themes in tech. If you would like to receive it directly in your inbox, subscribe now.
In this edition of Request For Product (RFP), I’ll provide an idea dump of sports- and sports betting-related products that I find thought-provoking. The genesis of these ideas came from a combination of pain points I’ve encountered within the space and recent conversations I’ve had with industry founders and analysts. As a reminder, this format is intentionally high-level, focusing more on ideation and inspiration than on granular analysis. At the end of the piece, I’ve included a simple poll to let readers vote on this edition’s ideas.
💬 Sports betting conversational media
Digital interactivity is a keystone of sports betting. Think about the difference in engagement between watching a game in which you have no action versus a game in which you have a bet on the line. The latter is filled with additional interactions across your digital networks, trading banter and discussion with friends and strangers alike. These digital campfires occur across a variety of platforms, from group texts to heated online forums to Twitter.
Select companies and startups have started to consolidate the discussion fragmentation, creating endemic social ecosystems within native applications such as theScore, Betsperts, Bookit Sports and Chalkboard. Even DraftKings has thrown its hat in the ring, recently announcing that it will roll out social features across both its sportsbook and DFS products to facilitate peer-to-peer information dissemination and to keep users within the digital walls of its app.
At the foundation of all social app experiences are text-based interactions, whether it’s via messaging or posting. Recently, additional content layers that sit on top of text-based interactions have emerged, enriching conversations while adding nuance, personality and culture. The tech world calls this format ‘conversational media’. You’re probably most familiar with this format when you pay a friend on Venmo and the text input box recommends an animated sticker.
This Venmo feature is currently powered by Holler, the market leader for industry agnostic conversational media creation and distribution. In a recent report, Holler detailed how consumers feel about enhanced messaging content in a digitally driven world1:
70% of people would share branded content they thought was funny or cool
78% believe visuals convey what words can’t
70% want more variety in visuals
Given the rapid growth of sports betting in the US and the uptick in social betting experiences, there’s a white space opportunity for an endemic conversational media provider to exist. A sports-exclusive approach would afford dev resources the dedicated focus required to generate animated stickers based on live developments pre- or post-game (e.g., dynamic odds, bad beats, crazy plays, meme-able moments), creating a treadmill of contextually-relevant, fresh content for users to include within their communications. This approach is similar to a Live Ops team in gaming that’s responsible for continuously updating content and live events to keep users continually engaged and retained. However, it’s important to note that large game publishers benefit from having deep relationships with the app stores, affording them app store approval expedience. For a newer or smaller company, the lack of relationship or timeliness could affect the contextual relevance of deployed content and is therefore an operational risk to be aware of.
In terms of branding, conversational media represents a new channel for operators to become part of the conversation. Imagine a DraftKings branded pre-game sticker depicting the current Total for a game with an animated team-themed hammer smashing the number. Even if dynamic odds aren’t technically feasible, the smashing of the word ‘Over’ would still suffice with the help of some nifty design. For reference, Holler ran a campaign with Chipotle in 2020 in which the animations were shared by over 129K people and viewed 22.4M times, reaching 4.3M consumers and becoming part of 150K conversations.
Whether it’s a native app that integrates animated stickers across WhatsApp and iMessage (i.e., B2C) or an SDK that allows any of the aforementioned social sports betting apps to enhance in-app conversations (i.e., B2B), sports betting conversational media can further enhance digital interactivity as legalized betting expands - not to mention the emergence of esports betting and its digitally native audience who view this functionality as table stakes.
🤓 Data scientist marketplace for sports analytics
The ability to provide storylines through data is the lifeblood of sports. While media companies and sportsbooks have the most obvious and commercially-relevant use cases for data, everyday fans also leverage data to power important facets of their lives. This includes individual creators seeking to enrich their content, academics in need of supporting data for a thesis or bettors looking to build a predictive model. However, accessibility to core sports data stands as the biggest barrier to most stakeholders. They either don’t have the technical skills required to acquire and analyze large datasets or do, but are deterred by the hurdles of maintaining multiple web scrapers and cleaning data from disparate sources.
All of the aforementioned stakeholders would benefit from the existence of a two-sided marketplace that facilitates connecting sports-focused data seekers with sports-focused data scientists. Data seekers could browse and purchase existing analyses on the platform or request a bespoke analysis to be completed on a project basis.
Additionally, this newly created ‘sports data scientist for hire’ platform could be created in partnership with a major sports data provider such as Sportradar, who would provide data accessibility and benefit from the bottoms-up exposure. The product experience would be similar to a platform like UpWork, but with the added layer of Sportradar’s core stats APIs. In other words, the hired data scientist would have direct access to a breadth of verified data to work with, streamlining the completion time for each job and creating a top of funnel lead for Sportradar should the client or data scientist have evolving business aspirations. In terms of economics, Sportradar would assumedly share in the take rate that the platform charges
You may be asking yourself “Wait, why can’t this use case be serviced on Upwork?”. The answer is that you can use UpWork, but it still comes with the inherent issue of data fragmentation and the potential lack of endemic knowledge that could complicate communication and compromise the quality of the end product. Anecdotally, I’ve been there before - it’s not fun.
Building a platform that not only allows statheads to monetize their sports data wrangling skills, but also provides an endemic analytics resource for sports data seekers creates a passion-driven marketplace with demand that grows alongside sports fandom.
👂 Asynchronous audio for sports betting
2021 should be dubbed the ‘year of audio’ given the rapid progression of audio-based products amongst both startups and major social platforms. Let’s take a look at some of the biggest deals and developments within audio this year:
Synchronous audio (real-time):
Drop-in live audio app Clubhouse raised over $100M across two separate funding rounds since January. The latest round valued the company at $4B
Spotify announced that it has acquired Betty Labs, the creators of Locker Room, a live audio app that’s changing the way insiders and fans talk about sports. The deal was valued around $50M
Facebook launched ‘Live Audio Rooms’ and has also started testing a product allowing creators to interact live with their audience
Twitter launched its live audio product, Spaces
Reddit launched its live audio product, Reddit Talk
Asynchronous audio (not real-time):
Yac, an Orlando, Florida-based startup that’s digitizing voice notes for remote offices, raised $7.5 million in a new round of funding
Facebook announced a product that will let users record brief voice notes and post them in their News Feeds
Bravo to Clubhouse for catalyzing this broad push into audio. While most of the product rollouts above are within synchronous audio (real-time), there are equally as compelling use cases within asynchronous audio (not real-time). This is typically delivered via voice note format, allowing users to consume short-form content at their convenience while sustaining the intimate and immersive format inherent to audio.
In terms of a sports betting use case, there could be an effective short-form audio modality that lives somewhere in-between a Tweet containing a capper’s picks (i.e., not enough context in terms of the why) and an hour long betting podcast (i.e., too much context in terms of the what and why).
The format would be similar to breaking down an extensive betting podcast into on-demand voice notes mapped to each game/team/player, allowing a user to consume the information in an à la carte fashion. The product experience could be delivered via native app, utilizing a follower-based design in which one could follow both their friends and influential personalities. For example, imagine following your favorite Twitter capper on such a platform and being able to actually hear why they like a certain play in a quick bites format. That sure beats blindly following a picks slate that lacks any meaningful explanation. On the creator side, the level of effort is minimal; they just have to speak to the approach.
Sure, some of the old guard in sports betting will balk at this approach, expressing aversion to disclosing their rationale as it erodes their edge the next time that situation occurs. However, the world is changing. Digital natives are utilizing sports betting as an additional interactivity layer rather than a potential ROI play. When designing for scale in this space, it’s best to consider the consumer behavior of emerging generations.
The more I detail this idea in writing, the more I start to realize that it’s better suited as a feature (within a product like The Action Network or Betsperts) rather than a product. Regardless of which side makes more sense, the functionality certainly could provide a new consumption format that serves various use cases within consumer sports betting apps. For kicks, here’s a mockup of how it could function within the Action Network app:
🏟 F2P in-stadium games
One lever that sportsbook operators have been pulling to capitalize on the customer acquisition (CA) gold rush in the US is strategically acquiring or partnering with free-to-play (F2P) platforms and infrastructure. These F2P experiences provide a risk-free, accessible way for users to familiarize themselves with betting concepts and to indicate that they’re interested in predictive gaming. Ultimately, a conversion to real-money sportsbook activities is the end goal. F2P experiences also serve as an important tool for expanding existing player databases (i.e., CRM) and for retaining or reactivating existing users.
Outright acquisition of a F2P platform can provide operators with increased CA efficiencies via vertical integration while F2P partnerships provide turnkey solutions to shepard new bettors into operators’ acquisition funnels. Some operators have even chosen to build such experiences in-house.
Here are a few examples of relevant F2P-related acquisitions, partnerships and in-house products over the past year:
Acquisition
Partnership
Proprietary Products
FoxBet’s Super 6 (4.5M users!)
With fans expected to return to stadiums this Fall, I’d love to see a F2P prediction game built at the team level that offers fans a chance to refund the face amount of their ticket. The game design could consist of 10 (or so) novice-level questions relating to pre- and in-game outcomes. For example, ‘which team will win?’, ‘will player A throw a touchdown pass?’, ‘will there be a field goal in the game?’, ‘will team B score over 7 points in the 4th quarter?’. The number of correctly answered questions results in promos from sponsors, with the grand prize of a full ticket refund if a user answers all 10 questions correctly. This type of experience not only affords a sportsbook with localized targeting and access to new demographics, but also affords a team organization with incremental, indirect stadium revenue.
A team-based approach grants operators access to a prospective regional bettor pool that may otherwise be uninterested in playing predictive games that require a user to make predictions across the entire league. In this format, predictions only pertain to a particular team and game that a user (or superfan) may have more localized knowledge about, making participation more appealing. Additionally, practically speaking, who wouldn’t answer 10 quick questions if it gives them a chance to win back the $300+ they sunk to attend the game. With atypical betting prospects participating in the contest, some may realize that they actually enjoy the experience of predictive gaming, representing an expansion into a new demographic and potential real-money conversions down the road. And let’s not forget that the average per game attendance for the NFL is 66K. Across 16 games, that’s 1M fans per week to target. Sure, we’d need to exclude all fans under the age of 18, but that TAM is still a big number compounded across the season and across additional sports leagues.
Team organizations stand to benefit from this F2P experience assuming fans stay in seats longer to see how their predictions unfold. For example, some questions could be structured around a late-occurring event (e.g., ‘will team B score over 7 points in the 4th quarter?’), driving fans to stick around to see the outcome. Increased time spent in the stadium results in a higher chance of ancillary purchases (e.g., concessions, merchandise). Additionally, the venue operators could make an experience out of it, displaying the list of remaining fans with correct picks on the jumbotron as the game unfolds (kind of like HQ…RIP).
So how would this be facilitated? I see two options. An operator could spin this program up, building and shipping the app experience and being responsible for the ticket payouts. Let’s use napkin math to quantify the economics. For one NFL game, we’ll use the previous data point of 66K total attendees. Let’s say that we get 25% of the attendees to engage in the F2P prediction game, or 16.5K participants. Correctly predicting 10 events on an all-or-nothing basis is essentially a 10 game parlay. The fractional odds for a 10 game parlay bet, based on -110 prices for example sake, is 720 to 1, or an implied probability of 0.14%. Multiply that by the 16.5K participants and that leaves us with 23 participant ticket payouts. At an average ticket price of $300, that amounts to $6,900. On a unit basis, this means an operator would have paid 42 cents for each of the 16.5K new database entries in their CRM ($6,900/16.5K). Additionally, F2P leader Chalkline cites that 20% of all F2P players click on sportsbook revenue activities. If we then make a conservative assumption and say that 3% end up becoming first-time depositors (FTD), that results in nearly 500 new players for the sportsbook (16.5K * 3%), or a ~$14 CAC ($6,900/500). Not bad compared to the $150-400 CACs operators are realizing organically or via affiliates.
Alternatively, a company could offer a white-labeled software solution to teams that provides all of the necessary components such as the game UX and question and data generation. Revenue would be driven by SaaS licensing to teams and exclusive sponsor fees or affiliate rev share from sportsbook partners. Authenticating and verifying ticket price would certainly be a problem to solve for in both models (good use case for blockchain-based ticketing), but the overall design warrants some food for thought.
Poll: Which idea was your favorite?
💬 Sports betting conversational media
🤓 Sports data scientist for hire marketplace
👂 Asynchronous audio for sports betting
https://lp.holler.io/final-lp-messaging-report