Platform selection: one week to a written decision

TL;DR Platform selection is one week of work and one written decision memo, priced at $2,500 and credited to the build when you go ahead. Five decisions, one recommendation, a reason recorded beside each, and the trigger that would reopen it later.

You are not buying a comparison of brands. You are buying the moment the platform argument inside your own building ends, because the answer arrives written down, with a price beside it and a reason that can be checked. The week settles the question before the money is spent, not after.

The choice of platform is a cost you pay once. Most stack consulting stops at a recommendation; this week stops at a signed page, and the signature is what turns it into a decision. We work in English and in Spanish, from Brickell, on site across South Florida.

What this includes

• A written decision memo, one page per decision, in the language your team works in, dated and signed at the bottom.
• A shortlist of two or three candidate platforms, each priced across year one and year three, including the option we would not be paid to build.
• A named owner of the data, recorded in the memo with the hosting account, the domain and the billing account.
• A performance target stated in plain words and tested on a real phone on a Miami mobile connection.
• An explicit list of what we are not going to touch, agreed in writing before any work starts.
• A reversal trigger beside each decision, so the memo says in advance what would reopen it.
• The week itself, run inside the consulting practice this week sits inside and starting from measuring the store you already run.

The decisions the memo has to settle

The first decision is the storefront your customer sees: the page, the photographs, the size of the buttons and the time the first screen takes to appear on a phone held in a parking lot in Doral. The second is the engine that holds your products and your orders, the piece that keeps stock counts honest, takes the payment and hands the order to whoever packs it. Those two are often the same piece of software and they do not have to be, and the memo states which one you are choosing and why, in a sentence a shop owner can read out loud. This is not a general tech stack consulting survey that arrives as a slide deck. It is five decisions, one page each, with a name and a date underneath.

Data is the third decision and the one most often left blank. Where does the customer record live, who can export it without asking permission, and whose card pays the bill when the platform changes its prices? We write the name of a person, not a department, beside the data, and we move the accounts into your name before the work starts. The fourth decision is what has to be true in real time on your busiest day: how fast a page has to open on a phone, how quickly a stock count has to correct itself when two people buy the last item in the same minute, and what your team needs to see on a screen when orders arrive faster than the printer can print them. the layer every store already pays for is usually part of that answer, and the memo explains in plain words what it does for you. If speed is the number you care about, the speed numbers Google publishes are the ones we hold ourselves to. Both targets are measured the same way: on a phone on a mobile network in South Florida, never on a laptop on office wifi, because a phone in a customer's hand is where the order is either placed or abandoned.

The fifth decision is the shortest and the most useful: the list of what we are not going to touch. If the store already ranks for addresses that bring orders, those addresses stay exactly as they are, and the memo says so before anyone opens an editor. Beside each of the five decisions we add a reversal trigger, the one event that would make us reopen it. A platform that stops letting you export your own product list is a trigger. A hosting bill that grows with no change in traffic is a trigger. A store that outgrows its own checkout on the busiest weekend of the year is a trigger. Naming them in advance is how a written decision stays a decision instead of drifting back into an argument two months later.

How it works

1. Discovery. One week, $2,500. We read the store as it stands, talk to whoever runs it day to day, and write the memo. At the end of the week you have the five decisions and a recommendation, whether or not you continue with us.
2. Architecture. $2,500, credited in full to the build if you go ahead. This is the architecture review in written form, where the chosen platform becomes pages, data, accounts and a sequence of work.
3. Build, on the build side of the decision. From $5,000. We build the store the memo describes, and we do not add a feature the memo did not ask for.
4. The move, if you are already trading. On what moving a live store involves, we carry across the catalog, the orders and the addresses that bring traffic today, and we check each one afterwards.
5. Care. From $1,500 a month. Updates, monitoring, small changes and a person who answers when something breaks on a Saturday.

A memo is only worth the week it takes if somebody signs it. Ours carries the name of the owner on our side and the name of the owner on yours, and it is dated. The week also asks something of your side: roughly a day and a half of your team's attention in total, mostly from the person who runs the store day to day, plus a decision maker in the room at the end for the signing. That is the whole ask, and it is why the week fits inside a normal trading week instead of stopping the business for it.

If the answer turns out to be to change nothing, that is a written outcome too, and a good one. A store that is already fast, already owned and already paid for does not need a new platform, and saying so plainly is cheaper than any rebuild. A memo that says change nothing still names the owner of the data and still carries a reversal trigger for every decision, so the file stays useful the day a new price, a new owner or a new competitor changes the question. What follows the week is only ever the part that the written decision asked for.

What it costs

Prices below are where each phase starts; your scoped quote fixes the exact scope, the price and the timeline.

Discovery is $2,500 for one week. Architecture is $2,500, credited to the build. Build starts at $5,000. Care starts at $1,500 a month. The same list sits on the published price of every phase, and it is the list we quote from.

The figure people forget is the one that arrives every month whether or not anything is being built. what a hosted plan actually rents you is a line on your card for as long as the store is open, and the memo compares that line across year one and year three before you choose. Two options can look identical on the day you sign and then differ by a few thousand dollars a year later, and that difference is what decides which of the two is the cheaper store. Year one carries the setup, the migration and the new habits; year three carries the renewal, the raised limit and the hours your team no longer spends keeping it alive. That is why the week is priced as a week and not as a percentage of whatever you spend afterwards, which also keeps our advice honest: we are paid for the week, not for the size of the build that follows it.

Proof

Food By Box runs an online store carrying 2,500 products with a Google speed score of 95. Monthly visits grew from 20,000 to 500,000 in 24 months, and every address that was already bringing traffic kept after the work. Tatiana Miath was the harder test: on a phone, a page that took just under five seconds to appear now appears in barely over one second, and nothing the store ranked for was lost. the numbers behind these projects are published in full. Michael Agnassia writes the architecture and reads the code before it ships, after 15 years in e-commerce, and who signs the decision and checks the build is written on the memo before anything ships.

Questions

What does platform selection cost and how long does it take?

It costs $2,500 and takes one week. The week ends with a written decision memo, a recommendation and a reason beside every decision, and the $2,500 is credited to the build if you continue with us.

Does platform selection have to end in a change?

No. Changing nothing is a valid result of the week, and we will say so in writing if that is where the evidence points. A store that is already fast, already owned and already paid for is one you should probably leave alone.

Who owns the data after the week ends?

You do, and the memo names the person who holds it. The domain, the hosting account and the billing account sit in your name before any work starts, so leaving with everything is a matter of exporting, not of negotiating.

What happens if the platform we choose turns out to be wrong?

Every decision carries a reversal trigger, so the memo already states what would reopen it. If that trigger fires, we return to the written decision instead of to the original argument, and we change the one thing that needs changing.

How much of our team's time does the week take?

About a day and a half in total, spread across the week. We need the person who runs the store day to day, a look at the numbers you already have, and a decision maker in the room for the signing at the end.

Will you touch our live store during the week?

No. The week produces a document, not a change to your live store, and the store keeps trading exactly as it does now. Work on the store begins only after the memo is signed, and the list of what we will not touch still applies after that.

Next

The week is booked as a week and not as an open-ended engagement, and it starts on the Monday your calendar allows. By the Friday you have a signed memo, a named owner of the data and a decision you can defend with a document rather than a feeling. pick a week that suits your calendar and we will confirm the dates and the scope of that Discovery week.