The thing nobody tells you about automation projects is that most of them die before they go live.
Not because the technology was wrong. Not because the AI was not ready. They die because someone budgeted three months for what should have taken three weeks, the project sprawled, the brief drifted, two key stakeholders left, and by the time the slack channel went quiet the original problem had moved on.
We picked 14 days for a reason. Here is why, and what we had to cut to make it work.
The maths of momentum
A two-week build is not a constraint, it is a forcing function. Two weeks is short enough that you cannot afford a kickoff workshop, a discovery sprint, a stakeholder alignment session, a tooling RFP, or any of the other rituals that consultancies invented to defend their hourly rates.
Two weeks is also long enough to ship something real. We are not racing. We are removing the rituals that do not earn their place.
The shape of a typical CueBot build:
- Day 1: Kickoff call. We watch you do the work we are about to automate. No slides.
- Day 2-3: Access to the tools we need (Shopify admin, Klaviyo, Stripe, whatever). We start building.
- Day 4-7: First working prototype runs against real data, end-to-end. You see it actually fire.
- Day 8-11: Edge cases, error handling, the unglamorous bits. This is where most builds quietly fail.
- Day 12-13: You use it for a day. We watch the logs. We fix what breaks.
- Day 14: Handover. Documentation. You own everything.
Notice what is not in there: design reviews, pixel-perfect dashboards, training sessions, change management, internal communications strategy. These are not free. They are also not the work.
What we cut
To ship in 14 days, we cut three things every other agency includes.
No discovery deck. We do not write a 40-page document explaining your business back to you. If we cannot understand your business in the first 30 minutes, you are not the right fit and we tell you so on day one.
No bespoke admin UI. Every CueBot build runs on tools you already use. Shopify admin is your dashboard. Klaviyo is your email control panel. Stripe is your invoicing. We do not build you a portal because portals decay and yours will already exist in 18 months.
No phased rollout. Either it works on day 14 or we are not done. Phased launches sound responsible but they are just project managers stretching the timeline. We commit to the full scope, ship the full scope, and call it done.
Two weeks is short enough that you cannot afford rituals. It is long enough to ship something real. The point is not the speed. The point is the discipline.
What we keep
The discipline of 14 days bought us back a lot of things, but the most important one is honesty about scope.
We cannot oversell. If you ask for cart recovery, social content generation, customer support, blog SEO, and competitor monitoring all in one go, we will say so. Pick the most painful one. Build that in two weeks. Add the next one in another two. We have shipped clients five automations in eight weeks this way. Never six in two.
We also keep the pricing honest. Our service prices (£700-£1,200) and pack prices (£900-£4,900) assume a 14-day timeline. If we drag past it, we are losing money, not invoicing more. That alignment matters more than any contract clause.
What this means for you
If you are about to engage anyone (us or anyone else) for an AI build, ask them three questions:
- How long will this take?
- What is the success metric, and what happens if you miss it?
- After you ship, what stays in your control versus theirs?
The answers to these tell you whether you are buying software or buying meetings. If a vendor cannot answer the first one in days (not "it depends"), or the third one in plain English, walk away.
The 14-day rule is not really about speed. It is about whether the person opposite you has the discipline to commit to an outcome instead of an open-ended engagement. Speed is just the receipt for that discipline.
We get this wrong sometimes. There are builds that take 16 days. There have been a few 21-day jobs where the integration was harder than scoped. When we miss, we own it, we tell the client, and we eat the cost. That is also part of the contract.
What we do not do is dress up the slip as "phase two".