The MP4 is dead

The MP4 container format, born in 2001 as ISO/IEC 14496-1 and refined as MPEG-4 Part 14 in 2003, is dead. Not for movies. Not for streaming. Not for the social feed. For marketing communication that needs to reflect live customer data at the moment a person opens it, MP4 has reached the end of its useful life. The replacement is not a better codec. It is a fundamentally different architecture: live, code-based video that assembles on the recipient’s device. Blings calls it MP5.

This post is the obituary, the post-mortem, and the migration plan. We will cover what MP4 did well for 25 years, why it stopped serving marketing, what specifically killed it, and what comes next.

A short obituary

MP4 was the right answer to the question marketing was asking in 2001. Distribute one video to many people, encoded once, played back consistently across devices and platforms, with a small enough file to move over a 56K modem and a polished enough output to look correct on a 4:3 CRT. It made YouTube possible. It made Netflix possible. It made TikTok possible.

For a quarter century, it was the universal language of digital video. Every marketer, every email service provider, every CDN, every ad platform speaks MP4 fluently. That ubiquity is what made it dominant, and it is also what made the next thing so hard to see coming.

MP4 is survived by its broader family of MPEG standards, which continue to do excellent work in entertainment and social media, and by its grandchildren in the streaming and codec world. The cause of death, in marketing, was a mismatch between what the format can do and what marketing needs it to do.

Where MP4 served us well

Three things MP4 did better than anything else for two decades:

• Universal compatibility. An MP4 plays on every device, in every browser, in every email client (when supported), in every social feed. That portability is unmatched.
• Compression efficiency. Successive codec generations (H.264, then H.265, now AV1) made high-quality video feasible over imperfect networks. The MP4 container made it portable.
• Predictability. A 30-second MP4 is a 30-second MP4. The bytes do not change. The runtime is fixed. The compute happens once, on a server, by the producer. Operations teams love predictability, and MP4 delivers it.

This is the part of the eulogy where you say, with a straight face, that MP4 leaves the digital world a richer place than it found it. That is true.

Where MP4 stopped serving us

The needs of marketing communication diverged from the strengths of MP4 over the past five years. Three shifts changed the math.

1. Data became the creative. A welcome video that shows the customer’s name is one thing. A loyalty video that shows the customer’s current point balance, expiring offers, and next-best action is another thing entirely. The first can be pre-rendered. The second cannot, because the data shifts between send time and open time.
2. Personalization moved to one-to-one. A pre-rendered MP4 is a one-to-many asset. Marketing went one-to-one and discovered that pre-rendering one video per customer at scale is operationally absurd: storage, compute, re-renders, version control, expiration, all of it.
3. The Moment of Open became the only moment that matters. The most accurate data is the data that exists at the millisecond a customer opens the asset. MP4 cannot read that data. It was rendered hours, days, or weeks earlier. The data inside the file is, by definition, stale.

This is what we call Data Decay: personalization that was accurate when it shipped but inaccurate by the time it was viewed. MP4 is the format that makes Data Decay inevitable.

What specifically killed it

Three architectural realities made MP4 untenable for personalized marketing at scale, regardless of how good the encoder got.

• Server-side rendering. Producing one MP4 per customer requires a server farm. The bigger the audience, the bigger the farm. Costs scale linearly with sends. Many enterprise teams ration personalization to protect the rendering budget, which defeats the purpose.
• Static output. Once the bytes are encoded, the bytes are encoded. If the customer’s loyalty balance changes between Tuesday and Thursday, the Tuesday MP4 is wrong on Thursday. The only fix is a re-render and a re-send.
• PII transit. To produce a personalized MP4, customer data has to travel to the rendering vendor’s cloud, get composited into the video, and travel back. For regulated industries (banking, healthcare, telecom), every campaign becomes a security review, every time.

None of these are bugs in MP4. They are consequences of pre-rendering as a delivery model. MP4 was built to be encoded once and consumed many times. Marketing has moved to a world that needs the opposite: encoded never, assembled fresh at every view.

What replaces it: MP5 AKA smart video

The replacement is not a new codec. It is a different category of asset entirely: a code-based video format that assembles on the recipient’s device at the Moment of Open, pulling live data from the CRM at the instant the asset is viewed. Blings calls this MP5 technology or Smart Video. 

The architectural inversion is the whole point:

• Client-Side Rendering. The video is built on the recipient’s phone or laptop, not a Blings server. No PII ever leaves the client’s environment. Blings calls this Zero-Knowledge Architecture.
• Live-Sync Content. Because the asset is code, the data layer is read at view time. A loyalty balance that updates in Salesforce shows the new number the next time the customer opens the email, with no re-render and no re-send.
• Template-based economics. One Dynamic Master Template generates unlimited variants. Cost scales with templates and campaigns, not with every render. Sending a million personalized videos does not cost a thousand times more than sending a thousand.

The trade-off is honest. MP5 is for marketing assets where the personalization needs to be live, the data needs to stay private, and the volume needs to scale without a rendering bill. MP4 is still the right answer for movies, social posts, broadcast advertising, and any one-to-many asset where the content is the same for every viewer.

For a deeper architectural look at why this shift was inevitable, see the Blings piece on why architecture matters more than the algorithm.

What does smart video look like in practice?

Three customers running MP5 at scale, as published evidence rather than promise:

• McDonald’s used MP5 to power its McCoins loyalty campaigns. The case study reports a 4.2x sales lift over benchmark, 5x app opens, 8.3x ROI, 54 percent video completion, and 26 percent CTA click rate. Since 2021, McDonald’s has sent nearly 100 million MP5 videos through Blings.
• Habit Burger Grill ran a loyalty referral campaign using personalized MP5 video. The case study reports a 47 percent boost in loyalty signups, a 53 percent share rate (roughly 13 times the industry benchmark), and a 44 percent lift in referral conversion versus their average.
• Mifal Hapais used MP5 to let existing members share personalized videos with friends. The case study reports a 27 percent share rate among engaged members and a 20 percent signup rate from the second-tier audience.

None of these results are achievable with MP4 in the loop, because none of them tolerate Data Decay between send time and open time.

Mourning practices: what to stop doing in 2026

If MP4 is dead for personalized marketing, here are the habits to retire alongside it.

1. Stop quoting “render minutes” in your video budget. Render minutes are a 2010 metric for a 2010 problem. The 2026 metric is templates and live data sources.
2. Stop accepting “we’ll re-render and resend” as a fix. Re-rendering means the original send was wrong. Live content means the original send is right at every open.
3. Stop sending PII to rendering vendors. The security review is no longer a tax to be paid. It is a constraint to be designed around.

Stop building campaigns around one creative for the whole list. If the asset cannot vary per recipient, it is broadcast, not marketing.

How to migrate

You do not need to rebuild the entire video stack on day one. A pragmatic migration looks like this:

1. Pick the highest-value campaign first. Loyalty, win-back, or onboarding usually pay back the fastest, because the data is most volatile.
2. Replace the static MP4 with a Dynamic Master Template. One template, many variants, all reading from the existing CRM.
3. Distribute via Live URL. Instead of attaching a file or hosting an MP4, share a URL that generates the personalized video on the recipient’s device at the Moment of Open.
4. Measure revenue per send. The point of the migration is not “we use MP5 now.” It is the conversion lift, the retained PII control, and the killed re-render cycle.
5. Roll out a second campaign within 30 days. The infrastructure is reused; the second campaign costs a fraction of the first.

For a step-by-step on how a personalized video email campaign actually gets built, see Blings’ guide on personalized video emails.

Frequently asked questions

Why can’t a “smarter” MP4 solve this?

Because MP4 is fundamentally a static container. Once the bytes are encoded, the data inside cannot update. You can re-render the file, but that means producing a new MP4 every time the data changes, at scale, for every customer. That is the problem MP4 cannot solve. The “smarter MP4” you want is actually MP5. 

What is MP5?

MP5 is Smart Video, Blings’ code-based video format. Instead of pre-rendering a video file, MP5 sends instructions and assets to the recipient’s device. The video is assembled on the device at the Moment of Open, pulling live data from the CRM. The result is a video that reflects the customer’s current data every time it is viewed, with no re-render and no PII leaving the client’s environment.

Does MP5 work in regular email clients?

Yes. The recipient receives a Live URL that opens in any modern browser. The personalized video plays the same way a normal video would, except the content is generated at the moment of view rather than pre-rendered.