Start with Berg is the way in. What it gets aimed at depends on what is actually stuck.
The offerStart with Berg$6,000two weeks, one read and one fixWhere it applies
The cheapest, fastest way to work together. You get the diagnosis in writing and one of its fixes built inside two weeks, not a list you have to hire someone else to act on.
The first days go to reading the situation: the site, the numbers, the tools, and whatever your team is doing by hand. I come back with what I would fix first and why. We agree on one thing, I build it, and you get a short before-and-after so the change is visible rather than asserted.
One shape, one price, four possible outputs. On the fit call I tell you which of these your situation calls for.
No live site access required to start. A shared document or a staging link is enough.
These are not separate packages. They are what Start with Berg turns into, and the direction a bigger build would take from there.
Content and publishing
Articles and pages built from your own material, a refresh of the pages you already have so they stop sliding, and the search and AI-answer research that decides what is worth writing next.
The bigger build in this lane is the Content Engine, a system your team runs instead of pages ordered one at a time.
Lead systems
Capture and intake, enrichment so a lead arrives with context attached, scoring against the clients you actually want, routing to the right person, and the follow-up that closes the silence.
Bigger builds here look like a research system that finds and qualifies prospects on a schedule instead of waiting on a form.
Workflow and AI adoption
Automations and digests that replace manual checking, internal tools your team runs itself, and honest training on where AI helps and where it quietly makes the work worse.
Teams that want to run AI themselves usually follow with a working setup and the review habits that keep the output trustworthy.
Software and tools
Internal tools, assistants that sit where your team already works, calculators and assessments your buyers use, and prototypes that answer a question before anyone commits to a full build.
If the prototype earns it, the full build follows as its own scoped project with its own timeline.
Four honest options, and I will tell you which one I think you should take even when it is the first one.
If you have published good work and watched it go nowhere, the cause is often underneath the page rather than in it: crawl and indexing errors, missing schema, or two of your own pages competing for the same search. None of that is visible from the outside, which is how firms end up paying for the same content twice. Usually the repair folds into the work already scoped. If it is big enough to fill the two weeks by itself, I will tell you before you pay anything and price it separately.
Some teams take the keys the day training ends. Others want the work running at full speed while their own people ramp up. This is the second one: an extra marketer and builder without the headcount.
At the upper end it becomes strategy too: where to aim next, which channels to open, and tuning what is already live so more of the traffic converts.
It runs on a three-month first term, which is about how long steady work takes to show up in the results. By then you can see what moved and decide where to go from there.
Expert-services firms where the expertise is the product. Consultancies, brokerages, recruiting boutiques, research and data companies, and professional practices. Past clients have come in through all four doors: a brokerage that needed content, a software company that needed leads, a founder drowning in manual coordination, and firms that needed a tool nobody had built yet. If you are pre-funding and hoping to trade equity for work, this is not a fit.
That is the normal case, and it is the reason the two weeks starts with me looking rather than you choosing. Some version of all four is usually present. My job is to tell you which one is actually costing you money right now, because fixing the third-most-expensive problem first is how a year disappears. If two are tangled together, I will say that too and quote them as one piece of work rather than pretending they are separate.
I tell you, before I build anything. It happens. Someone comes in certain they need content and the real issue is that the leads they already have never get followed up. You are paying me to be right about that, not agreeable. We either point the two weeks at the thing I found, or you keep the read and we stop there.
A private publishing system built on your own material: voice research and a locked style guide, a custom agent skill that writes in your voice, a quality gate that runs on every draft, publishing wired into your CMS, and training so your team runs it without me. It also brings existing pages current rather than only writing new ones. Three to four weeks, quoted per project. It is where the content work ends up, not where it starts.
Not to start. For content I can hand over finished files your team publishes. For automations and lead work I can build against test data or a copy first. Access makes the loop faster and usually becomes part of the work once we are past the first two weeks, but a slow permissions conversation should not be the reason nothing gets built.
Start with Berg is $6,000, paid upfront, one invoice. Larger builds split in half, one payment at kickoff and one at handoff. Ongoing support is billed monthly against a three-month first term. Anything scoped after the two weeks gets quoted in writing before it starts, so nothing arrives as a surprise line item.
That can work, as long as I am not the third vendor in the room negotiating with the other two. The best version is when your team or your agency handles volume and I build the system or fix the thing they cannot. The version that fails is when nobody owns the decision and I am waiting on two approvals for every change.
A few things, and the reason is the same each time: you should end up with something working. So no standalone audits, code review, or technical due diligence unless a build is attached, because a document on its own still leaves you hiring someone else to act on it. Open-ended retainers go the same way, since every month should end with finished work you can point at. Social media management, video editing, photography, and customer interviews are all better in a specialist's hands than mine, and I can usually recommend someone.
I build with Cursor, Claude Code, and modern frameworks: Astro, Next.js, Supabase, Railway, and WordPress when that is where you already live. For automations, whatever your stack will connect to. The AI-native tooling is why two weeks of my work covers what an agency quotes for a quarter.
Two active builds at a time. That cap is what makes the timeline real, because when I take your project it has my full attention. When both slots are full I say so instead of booking work I cannot start yet.