Who builds Ralph, and
why does he exist?

The Friday afternoon that started it

We were running paid media for a fashion brand when a jacket's size 12 sold out at three minutes past two on a Friday. Nothing told us. The parent product still said in stock in Shopify, because the 6, the 8 and the 16 were sitting on the shelf, so the catalogue kept the product live and the two campaigns using it kept spending. A static ad we had shot that week showed the jacket on a model who was, of course, wearing a 12. By the time someone opened Ads Manager on Monday, roughly £180 a day had gone on clicks that landed on a greyed-out size, and the customers who did buy took the 16 and sent it back.

None of that was a bid problem or a creative problem, and the tools we paid for could not have caught it, because each of them could see either the warehouse or the ad account and never both. The stock report said what was low. The spend report said what was expensive. Nobody, and nothing, put the two next to each other on a Friday night. That join, size by size, is what we set out to build, and it is still the centre of the product.

What we had already learned from the agency side

Must Be Agency ran Shopify growth retainers for years before Ralph. We learned what merchants actually get for the money. A light retainer runs £750 to £2,000 a month in fees, a solid one £2,500 to £6,000, and Plus work starts at £6,000 and goes upwards, before any ad spend or creative production. We have written the bands up in what a Shopify retainer costs. A large part of every one of those retainers was the same overnight work: checking stock against spend, pulling hourly breakdowns, tidying titles so Shopping would accept them, and drafting the next campaign package for someone to say yes to.

That half is patient, repetitive and mostly reading. It does not need a senior strategist, and it does not happen at weekends because the strategist is asleep. The other half of the retainer, the taste, the brand calls and the hard decisions, is what people are worth paying for. We wanted software that did the first half every night so the second half got the attention.

What we believe

We believe you own the store, so anything that spends money or changes a live campaign waits for your approval, and that is how the product works rather than a line in the footer. We believe small shops count, and that a founder who is also the buyer, the merchandiser and the media buyer should not need a ten-person team to run a tight operation. We believe catalogue, ads, stock and demand belong in the same decision, not in four tools that never meet. And we believe in saying what the software does not do, in plain words, before anyone pays for it. There are no fake live counters on this site, no silent publishes, and no set-and-forget claims we would not run on our own stores.

Who is behind it

Ralph is built by Must Be Agency, and the team are practitioners who have run growth for real Shopify brands rather than a research lab shipping demos. If the product feels opinionated, that is deliberate. The opinions come from broken ROAS screenshots and late stockouts, not from a pitch template.

How we built him

Shopify is the spine, and he reads it at variant level, because a parent product can be half in stock while every size that sells is gone. Meta, Google, Klaviyo, Search Console and the weather feed in as context around that spine. Overnight, he works out days of cover per size and colour, finds the campaigns still paying for anything broken, pulls the hourly results and groups them into retail blocks, scans titles and attributes for what Shopping will reject, and puts true profit next to spend from your real orders. Everything that comes out is ranked by pounds at risk and drafted, not applied. The two mechanisms we are most often asked about are written up in full: how Ralph handles out-of-stock variants and how Ralph dayparts Meta and Google. The fashion design centre, where the size-run problem is sharpest, is in Ralph for fashion brands.

Then you say yes, or you do not. Approved changes are written with the reason recorded and a revert ready, and rejected ones simply close. The full operator walkthrough is in the docs, the safety model in trust, the category in what an AI growth operator is, and the longer answer to why the product exists in what problem Ralph solves.

Do it yourself first

Before we built Ralph we ran this routine by hand, and we would still tell any merchant to run it for a week before paying anyone, us included. On Monday, open Shopify Admin, go to Products, then Inventory, filter to zero available, and write down every size or colour that is out of stock while its parent is still live. On Tuesday, open Meta Ads Manager, set the last seven days, use Breakdown by product ID on each catalogue campaign, and write the daily spend next to each of Monday's rows; do the same in Google Ads under Shopping, then Products. On Wednesday, in Ads Manager choose Breakdown, then By time, then Time of day over sixty days, and write down the block of hours below the account average on most days. On Thursday, in Merchant Center open Products, then Needs attention, and list the disapprovals with their reasons. On Friday, apply the one fix with the most money behind it, note what you expect, and check on Monday. The total in your spend column is what the job is worth, and it is the number any operator, human or software, has to beat.

Where Ralph fits

Ralph does that week's work every night and hands you the list at seven in the morning with the fixes already drafted. Free sees everything and fixes 50 product titles and 50 Shopping attributes a month, with the writers switched off. Growth at £99 a month is the first plan where he applies what you approve; Pro at £249 covers three isolated stores and Scale at £599 covers ten. Founding members on the waitlist lock 25% off for life and annual billing gives two months free. The plans are in Ralph pricing explained, and how he sits next to ad automation, feed tools, attribution platforms and agencies is in Ralph vs the alternatives and Ralph vs an agency. Above ten stores we set things up by hand and talk like humans.

What Ralph will not do

He will not spend or publish without a yes, and every writer that touches money is off by default until you enable it. He will not write Google ad schedules, which are recommend-only today, and he will not let the weather move your hours. He only writes supplemental Merchant Center feeds and never overwrites your primary one. He will not invent a cost price to finish a profit chart, and he will not fix a bad creative or replace your taste. He is in private beta with a waitlist, billing is not open, and we would rather onboard operators carefully than inflate a signup counter. Every seat comes with a real person who walks you through the desk.

Who Ralph is for, and who he is not for

He is for the solo founder wearing every hat, the lean team without a media department, the multi-store operator who needs isolation and a clear morning, and the merchant tired of paying a retainer for overnight coordination. He is for anyone who wants AI help with a review queue in front of it. He is not for a team that wants unsupervised spend by default, for anyone shopping only for another attribution chart, for an enterprise that needs a custom data warehouse and nothing else, or for a store that will not connect Shopify, because without the store's own numbers there is nothing to optimise.

What we publish

We would rather ship operator guides and free tools than another list of the best software of the year. The pieces below are the ones we point new operators at first.

Questions we get asked

Get in touch

If you have a question about fit, multi-store setups or running Ralph alongside an agency, join the waitlist from the home page and say so in the note, or email rocketman@mustbeagency.com. If you already have a seat, sign in.