After spending years building go-to-market tooling and content workflows for B2B SaaS teams, I kept running into the same situation. A demand-gen team had a campaign to ship. They had a clear objective, an audience they understood, and a brief that was specific enough to get good copy out of a talented writer. And then the brief went to the agency.
Three weeks later, after two rounds of feedback, a direction change at the fourth stakeholder review, and a campaign that launched a month after the window it was designed for, the team had copy that was technically acceptable. It didn't sound wrong. It sounded like it could have been written for any company in the software space -- which, in a practical sense, meant it wasn't working for this company the way their best copy worked for them.
The root problem wasn't the agency's writing quality. It was the agency's inability to start from the brand's actual voice. Every external copywriter starts from a blank page and works toward your brand. Your best copy started from your brand and worked toward the blank page. That's a different process, and it produces a different result.
What We Built and Why
Draftlume started as an internal tool to close that gap. The idea was simple: if the bottleneck is getting a first draft that already sounds like the company, then the system should start from the company's materials -- the style guide, the past campaigns that worked, the vocabulary and argument structures that resonate with the specific audience -- and draft from there rather than from a general-purpose starting point.
We launched a private early-access program in early 2025 with a small group of demand-gen teams who had the same frustration we did. The consistent pattern we found was that the teams getting real results were the ones who had done the work of documenting their brand voice -- not perfectly, but specifically enough to give the system something concrete to start from. Teams who uploaded three or four strong past campaigns and a reasonably detailed style guide were generating first drafts that required editing rather than rebuilding. That's a meaningful difference in how the rest of the copy process runs.
The Core Mechanics
Draftlume ingests two things: your brand documentation (style guide, voice rules, banned phrases) and your past performance data (the campaigns that worked, with annotations about what made them work for which audiences). From those inputs, it generates first drafts for the content types your demand-gen team produces most: email sequences, landing page copy, LinkedIn ad copy, and short-form content.
The output isn't publish-ready without review. We've been explicit about this from the start because it's accurate and because teams that treat first drafts as publish-ready produce inconsistent content at scale. The value is in how far along the first draft is when a trained editor sees it. When the first draft is structurally correct for your brand and directionally accurate on voice, the editing work is refinement. When it isn't, the editing work is revision, which is a different (and much more expensive) activity.
For the teams in our early-access cohort, the median time from brief to first-draft approval dropped significantly. The more important change wasn't the time savings -- it was the change in what the team was doing during that time. Instead of producing copy, they were evaluating and improving copy. That's a more valuable use of a content strategist's attention.
What We're Building Toward
The current product solves the first-draft problem for teams that have done the work of documenting their brand voice. The next layer of the problem is making that documentation work accessible to teams that haven't done the documentation work yet -- and building the tooling to help teams extract a pattern library from their past campaigns systematically rather than manually.
We're also working on tighter integrations with the tools demand-gen teams already use. The most common request from our early-access cohort has been the ability to push a finished draft directly into HubSpot for scheduling rather than copying and reformatting manually. That's in progress.
Who This Is For
Draftlume is built for demand-gen teams that have a clear sense of their brand voice but struggle to transfer that voice reliably into high-volume copy production. If your team produces fewer than two campaigns per month and you have an in-house writer who knows the brand deeply, the tool will save some time but won't change your workflow substantially. If your team is producing four or more campaigns per month, relying on external resources for some portion of that production, and regularly running into the problem of output that doesn't sound like you -- that's the situation Draftlume is designed for.
If you've been waiting for the right moment to try building copy workflows around AI assistance, this is a reasonable starting point. We're bootstrapped, we're not optimizing for growth at any cost, and the teams using the product today are the ones who give us the feedback we need to make it better. Come try it and tell us what doesn't work. That's the feedback we find most useful.