Narrative Wants to Replace Your Entire Sports Highlights Team With AI and I Think Broadcasters Should Be Nervous

The Macro: Sports Highlights Are Still Made the Same Way They Were in 1998

I have been watching the sports media industry try to modernize for years now. The budgets are enormous. The timelines are absurd. And the fundamental process of turning a three-hour game into a two-minute highlight package has barely changed since SportsCenter was the only thing that mattered.

Here is how it still works at most broadcast operations. A game ends. An editor (or a team of editors) watches the footage, identifies key moments, cuts the clips, adds graphics and transitions, writes the voiceover script, records the audio, and publishes. For a single game, that process takes anywhere from 45 minutes to several hours depending on the sport, the outlet, and the stakes. For a full Saturday of college football or a packed Tuesday night in the NBA, you are talking about dozens of editors working simultaneously to keep up with demand.

The economics are brutal. ESPN employs thousands of people across its content operations. Regional sports networks that survive on highlight packages and recap shows run lean, which means fewer highlights, slower turnaround, and worse coverage of smaller-market teams. Minor league sports, college conferences outside the Power Five, and international leagues in smaller markets get almost no highlight coverage because nobody can justify the production cost.

Meanwhile, fans have moved to short-form platforms. TikTok, YouTube Shorts, Instagram Reels. They want highlights immediately after a play happens, not three hours later when the broadcast team finishes the package. The gap between what fans expect and what sports media can deliver has been widening for years.

WSC Sports has been the biggest player in automated highlights, working with leagues like the NBA and MLB to generate clips at scale. Grabyo handles live clipping for social distribution. But the full pipeline from raw footage to polished, personalized content package remains largely manual at most organizations. The tooling exists for parts of the workflow. Nobody has nailed the whole thing.

The Micro: AI That Watches the Game So Your Editors Do Not Have To

Narrative positions itself as the most advanced AI-driven sports content creation platform, and the pitch is straightforward: reach younger fans with personalized content at scale. The company came through Y Combinator and is targeting broadcasters, teams, and leagues directly.

The core value proposition is automating the creation of sports highlights. Instead of editors manually scrubbing through game footage to find the key plays, Narrative’s AI watches the footage, identifies the moments that matter, and generates highlight packages automatically. For teams and leagues that produce dozens of games per week across multiple venues, this is the difference between covering everything and covering almost nothing.

What makes the approach interesting is the personalization angle. Different fans want different things from the same game. A casual viewer wants the top five plays. A fantasy sports player wants every touch from their roster players. A team’s social media account wants hero moments for their stars. A betting platform wants plays that affected the spread. Traditionally, each of those outputs requires a separate edit. If the AI can generate all of them from the same source footage, the content multiplication is significant.

The target customers tell you where the money is. Broadcasters are spending millions on production teams that could be partially automated. Teams are trying to feed social media channels that demand constant content. Leagues want to increase engagement with younger demographics who consume sports primarily through clips, not full broadcasts.

The competitive dynamics are worth noting. WSC Sports raised over $100 million and works with major leagues. Pixellot does automated production for amateur and semi-pro sports. Grabyo focuses on live clipping and distribution. Narrative is entering a space with established players, which means the technology needs to be meaningfully better or meaningfully cheaper (or both) to win deals away from incumbents.

The Verdict

The sports content gap is real and getting worse. Every league wants more content across more platforms for more audience segments, and none of them want to hire proportionally more editors. Narrative is attacking a problem that is only getting bigger.

At 30 days, I want to see how the output quality compares to human-edited highlights. AI-generated sports content still feels robotic when the editing lacks narrative arc. A good highlight package tells a story. A bad one is just a sequence of clips. At 60 days, the question is customer adoption. Are broadcasters and teams actually replacing workflows, or is this another demo that looks great but does not survive contact with real production pipelines? At 90 days, I want to see whether the personalization engine delivers measurably different content for different audience segments, because that is the feature that separates Narrative from a simple auto-clipper.

Sports media is one of those industries that everyone knows needs to be disrupted but nobody has fully cracked. The last five years of AI progress in video understanding make this moment different from previous attempts. If Narrative can produce highlights that are 85 percent as good as a human editor at 10 percent of the cost and 100 times the speed, most sports organizations will take that trade. The question is whether the AI is actually there yet.