Three years ago I made a bet that video would become the default in B2B sales. I was right about the direction. I was wrong about the hard part. I spent the first two years worrying about the wrong bottleneck, and 2026 is the year that became impossible to ignore. The hard part was never getting reps to record. Recording friction is solved. The hard part is rewiring the default so video is the first thing a rep reaches for, not the special occasion they save for the top five accounts.
That wrong assumption cost me about eighteen months. I want to unwind it here because I think a lot of operators are sitting in the same wrong assumption right now, buying the tool and waiting for the default to follow. It does not. The default is a management habit, and management habits do not ship with the license.
This thread runs back through almost everything I have written at dearmer.com.au, but the cleanest framing comes from the first essay on AI as a management problem. The technology arriving does not install the behavior. The management work installs the behavior. Video as the default is the same shape of problem, and I was slow to see it because I was living inside it.
What I got right about video in B2B sales
The directional bet was correct, and the evidence finally caught up with it this year. Video is moving from the optional upgrade to the default reach for a growing share of B2B sales teams, and the shift is being driven by two forces that are now visible in the data rather than just in conviction.
The first force is the collapse of text as a signal. AI tools made competent B2B prose the floor, not the ceiling. Every cold email sequence sounds identical now because the same large language models produce the same grammatical patterns for every team running the same tools. When the noise floor is AI-quality text, AI-quality text stops being a signal. It is just more noise at a higher production value. I traced the early signs of this in the essay on video becoming the signal when emails look AI-written, and the trend it described has since become the default condition for most outbound teams.
The second force is the economics of personalization at scale. This is the part I did not see coming as fast as it arrived. When one recording can be cloned into thousands of individually addressed videos by name, company, and context, the marginal cost per personalized video drops to near zero. The cost argument against making video the default disappears with it. I have been writing about video as the B2B default long enough to remember when the objection was “it is too expensive to do at volume.” That objection is gone. What replaced it is harder.
The corroborating data point that changed my thinking this year came from outside Sendspark. Dan Rosenthal’s analysis of more than 250 go-to-market implementations, published at Growth Unhinged in 2026, documents that account-based teams running fifty-thousand-dollar-plus ACV now run a two-minute video as the demand-gen CTA that converts better than a demo booking. The outbound call to action is no longer “book a demo.” It is “would you like a two-minute video breaking this down?” That is not a video-vendor talking point. That is operators reporting what is winning in their funnels.
The wrong bottleneck I spent two years solving
I spent the first two years convinced the bottleneck was recording friction. If we could just make recording fast enough, low-friction enough, integrated enough into the tools reps already lived in, the default would follow. That assumption was wrong, and it was wrong in a specific, embarrassing way.
Recording friction is solved. It was largely solved by 2024. A rep can open a camera, record sixty seconds, and send faster than they can write a thoughtful email. The tooling exists, the integrations exist, the one-click record exists. And yet, in team after team, the default did not move. Reps bought the tool, did the training, sent a few videos in the first week, and then drifted back to text. The recording friction I was solving was not the bottleneck. The bottleneck was upstream of the recording.
The actual bottleneck is the cultural default. The default is what a rep reaches for first when a new name appears in the CRM and the clock is ticking. Under time pressure, every rep reverts to the format they already know. Text is the format they already know. Video is the special occasion they save for the accounts they have already decided matter. So the tool sits next to an unused default, and the team keeps emailing.
I should have seen this sooner. The pattern is identical to the one I wrote about in the founder insertion problem: the behavior you want does not install itself just because the capability exists. You have to redefine the default, protect the time for it, and inspect it. The same is true here. My background is in operating roles where the cost of a misread default was a missed quarter, and I still underweighted how much of video adoption is a management problem rather than a tooling problem.
The embarrassing part is that I had the frame already. The first essay on this site argued that AI is a management problem, not a capability problem. The exact same logic applies to video as the default, and I was too close to the product to apply my own frame to my own rollout. The technology arriving does not install the behavior. The management work installs the behavior. I spent two years solving for the technology and not enough months solving for the management.
Why personalization at scale collapsed the special-occasion framing
The special-occasion framing was always an economic argument in disguise. Video was the special occasion because it was expensive per prospect. A rep spent a full manual minute per video, which meant video was reserved for the accounts that justified the minute. That economics is what made video feel like an upgrade rather than a default.
Personalization at scale removed the economic argument. When one recording can be cloned into thousands of individually addressed videos by name, company, and context, the marginal cost per personalized video drops to near zero. The minute-per-prospect cost that justified the special-occasion framing is gone. What is left is the cultural habit that was built on top of that economics, still running after the economics underneath it was removed.
This is the part that operators miss. The habit outlives the reason for the habit. Teams that set their video policy in 2023, when each video really did cost a manual minute, are still running that policy in 2026 when the cost is gone. The policy feels rational because it was rational. It is just no longer attached to the economics that made it rational. Faithe Day’s analysis at Selling Signals on the AIO search shift makes the parallel point: as AI flattens written content into a sea of sameness, first-hand human proof on camera is the differentiator AI cannot fake. The economics of the special occasion are gone. The differentiator value of the human recording is higher than ever. The only thing still in place is the habit.
The teams that noticed the economics changed are the ones running the two-minute video as the demand-gen CTA. The teams that did not notice are still saving video for the top five accounts and wondering why their reply rates look like everyone else’s. The difference is not the tool. Both teams have the tool. The difference is that one team updated its read on what the tool is for, and the other is running a 2023 policy on 2026 economics.
| Framing | 2023 economics | 2026 economics | What teams still do |
|---|---|---|---|
| Cost per personalized video | One manual minute per prospect | Near zero, cloned at scale | Budget video like it still costs the minute |
| When to send video | Top five accounts only | First reach on any qualified prospect | Save video for the special occasion |
| Default first reach | Text email | Personalized video | Reach for text under time pressure |
| Objection to default video | ”Too expensive at volume" | "Reps will not do it” | Keep the old objection after the cost leaves |
| What the video proves | Extra effort on a named account | A human chose to show up for this person | Treat the proof as a luxury, not the signal |
The eighteen-month lag between buying the tool and moving the default
There is a lag between buying the tool and moving the default, and in the teams I have watched it runs roughly eighteen months. The lag has a shape, and recognizing the shape is the first step to shortening it.
The first quarter is enthusiasm. The team buys the tool, runs the training, and sends a burst of videos in the first week. Metrics look promising. A few replies come back that are visibly better than the text baseline. Everyone agrees this is the future. The dashboard gets shared in the Monday meeting.
The second quarter is drift. The burst fades. Reps revert to text under time pressure because text is the format they know and the default was never redefined. The videos that do get sent are concentrated on the same top five accounts. The dashboard stops getting shared. The tool is still installed, the license is still paid, and the behavior is back to where it started.
The third quarter is the management intervention, or it is not. This is the fork. Either someone on the leadership team redefines the default explicitly, protects recording time on the calendar, and starts inspecting the first video each rep sends each week, or the drift becomes permanent. The teams that compound take the intervention. The teams that stall do not. Harvard Business Review’s research on digital tools in B2B sales found that sales forces spend roughly a third of their time on non-selling activity, and that adoption discipline, not tool access, determines which teams convert digital capability into advantage. That finding applies here word for word.
The fourth through sixth quarters are compounding, for the teams that took the intervention. The default holds because it is inspected. Reply rates climb because the first reach is now the format that still carries signal. The reps who pushed back hardest in quarter two become the strongest advocates in quarter six, because they are the ones whose numbers moved. This is the compounding arc I described in the first essay, and it applies as cleanly to a video-first default as it does to AI rollout. McKinsey’s research on AI adoption gaps in 2025 showed roughly this pattern in early enterprise pilots: the teams that built the management discipline pulled ahead, and the teams that waited for the technology to improve did not.
The eighteen-month number is not a rule. It is a pattern I have watched across enough teams to take seriously. Your lag may be shorter if you run the intervention in week one instead of quarter three. It may be permanent if you never run it.
What a video-first default actually looks like day to day
A video-first default is a concrete operating choice, not a slogan. It shows up in what a rep does in the first thirty seconds after a new name lands in the CRM, and it shows up in what a manager inspects on Monday morning.
The default is this: the first reach to a new prospect is a short personalized video, not a text email. The rep opens their camera, references something specific to the prospect, and sends. Text becomes the follow-up format, not the first one. The default is defined by what the rep reaches for first when the clock is ticking and the quota is visible.
The video itself is short. Sixty to ninety seconds. The first ten seconds carry the specificity: the prospect’s name, a reference to something real about their situation, a reason this rep is reaching out now. The rest is the follow-through. Production quality matters less than teams expect. A slightly imperfect video with a genuine specific reference outperforms a polished video with a generic one, because the specificity is the signal and the polish is just tool access.
The management inspection is what holds the default. Once a week, the manager reviews the first video each rep sent that week. Not every video. The first one. The review is fifteen minutes per rep and it is the single intervention that separates teams that compound from teams that drift. You are not micromanaging the video. You are inspecting the default. The questions are simple. Did the rep reach for video first. Was the first ten seconds specific. Did the reference survive a follow-up question. That is the entire review.
This is the same management discipline I argued for in the handoff essay. The format that survives is the format the next person actually consumes. The default that holds is the default the manager actually inspects. Video as the default is not a tool decision. It is a management decision that happens to use a tool.
The window, and what outlasts it
The window for human video as a signal is open in 2026 and it will close eventually for the same reason text closed. When AI-generated video becomes genuinely indistinguishable from human recordings at near-zero per-prospect cost, the effort signal collapses. I wrote about that lag in the earlier essay on video as the signal: the window stays open longer than people assume, not because generation quality lags, but because detection calibrates faster than generation improves cheaply.
In 2026, AI-generated video still reads as slightly wrong in ways buyers can spot. The voice is a fraction off. The eye contact does not land. The micro-expressions are legibly synthetic. When buyers spot it, AI video is worse than text because it reads as elaborate deception rather than efficient volume outreach. Human video is still the signal. The window is real, and it is the reason the default is worth installing now rather than waiting.
But here is the part I want to be honest about, because it is the part I got wrong before. The default you install outlasts the format advantage. The discipline of researching a prospect thoroughly enough to say something specific in sixty seconds carries forward to whatever format earns the signal next. The teams that build the video-first default are not betting on video forever. They are building the research-and-show-up muscle that will outlast any single format. That is the compounding asset. The format is the vehicle. The muscle is the moat.
This is the lesson I keep relearning. I was right about video as the direction. I was wrong that the hard part was the recording. The hard part is, and has always been, the management work of rewiring a default under time pressure. The teams that do that work in 2026 will compound for the next eighteen months on a signal their competitors are still treating as a special occasion. The AI 2027 forecast describes the eighteen-month arc in the context of AI rollout, and the shape is the same: the teams that build the discipline early pull ahead in a way that becomes uncatchable, and the teams that wait for the technology to improve do not.
The closing thought
Being right about the direction does not mean being right about the timeline or the bottleneck. I was right that video would become the B2B default. I was wrong that the hard part was getting reps to record. The hard part is rewiring the default so video is the first reach, and that is management work, not tooling work. It is the same lesson the first essay on this site made about AI. The technology arriving does not install the behavior. The management work installs the behavior.
The economics that made video a special occasion are gone. The differentiator value of a human showing up on camera for a specific person is higher than ever. The only thing still standing between those two facts is a cultural habit running on 2023 economics in a 2026 market. The teams that notice the economics changed, and install the management discipline to hold the new default, will compound. The teams that keep saving video for the top five accounts will keep getting the reply rates of a team that still emails.
What is the first reach your team defaults to when a new prospect lands in the CRM, and when did you last inspect it?
Frequently asked questions
What is the hardest part of making video the default in B2B sales?
The hard part is not getting reps to record. Recording friction is solved. The hard part is rewiring the team default so video is the first thing a rep reaches for on a new prospect, not the special-occasion upgrade they save for the top five accounts. That is a management habit, and management habits lag tool purchases by roughly eighteen months.
Why did personalization at scale change the economics of video outreach?
Before AI personalization, each video cost a rep a full manual minute per prospect, which is why video was a special occasion. When one recording can be cloned into thousands of individually addressed videos by name, company, and context, the marginal cost per personalized video drops to near zero. The cost argument against making video the default disappears.
How long does it take a sales team to adopt video as the default?
Roughly eighteen months from tool purchase to cultural default, in the teams I have watched. The first quarter is enthusiasm, the second is drift back to text, the third is a management intervention that redefines the default, and the fourth through sixth are compounding. Teams without the management intervention stall in the drift phase indefinitely.
What is a video-first sales default in practice?
A video-first default means the first reach to a new prospect is a short personalized video, not a text email. The rep opens their camera, references something specific to the prospect, and sends. Text becomes the follow-up format, not the first one. The default is defined by what the rep reaches for first when the CRM shows a new name.
How do you install a video-first default without micromanaging every rep?
Define the default explicitly, protect recording time on the calendar, and review the first video each rep sends each week. The review is fifteen minutes per rep and it is the intervention that separates teams that compound from teams that drift. You are not micromanaging the video. You are inspecting the default.
Does video as the default still work when AI can generate video?
In 2026, AI-generated video still reads as slightly wrong in ways buyers can spot. The voice, the eye contact, and the micro-expressions do not land. Human video is still the signal. The window will close eventually for the same reason text closed, but the discipline of researching a prospect enough to say something specific in sixty seconds outlasts the format advantage.
Why do teams buy video tools and still not use them as the default?
Because buying a tool changes capability but not habit. The default is what a rep reaches for first under time pressure, and time pressure pulls reps back to the format they already know. Without a management intervention that redefines the default and inspects it weekly, the tool sits next to an unused default and the team keeps emailing.
Sources & references
- Growth Unhinged — How to build a modern ABM engine · Dan Rosenthal's 2026 analysis of more than 250 go-to-market implementations documenting that a two-minute video is now the demand-gen CTA that converts better than a demo booking at fifty-thousand-dollar-plus ACV.
- Selling Signals — How AIO Search Is Changing the Game for B2B Teams · Faithe Day's 2026 analysis of the AIO search shift, including the point that as AI flattens written content into a sea of sameness, first-hand human proof on camera is the differentiator AI cannot fake.
- Harvard Business Review — Integrating Digital Tools into Every Stage of Your Sales Strategy · Research showing sales forces spend roughly a third of their time on non-selling activity and that adoption discipline, not tool access, determines which teams convert digital capability into advantage.
- McKinsey — The state of AI in 2025 · Enterprise research on AI adoption gaps in 2025, documenting the early-stage compounding phase that separates teams who build the management discipline from those waiting for the technology to improve.
- AI 2027 Forecast · The research-backed forecast that introduced the scatterbrained-employee framing and the eighteen-month compounding arc that applies as cleanly to a video-first default as it does to AI rollout.