Confidential · Prepared for the Milton-Lloyd directors

Where AI actually belongs in Milton-Lloyd

Before you implement a single tool, you map the whole business — from the perfume oil arriving to the bottle landing on a customer’s doorstep — and find exactly where AI saves money, where it’s worth building, and where it’s dangerous.

A structured operational audit, tailored to a multi-channel fragrance manufacturer.

Made in England · Own e-com · Amazon · global seller network · 15+ audits delivered · 0 refunds

AI is mostly noise. The audit cuts it.

Every week there’s a new “revolutionary” AI tool. For a business owner that means one of three things happens — and none of them are good.

Either it’s so overwhelming you never start. Or you start, but you don’t know where, so it goes nowhere. Or you implement the hyped tool everyone was talking about — and it fails, because it was never right for your business in the first place.

The principle underneath all of it is simple: any process done on a computer will be 40–80% faster within the next two years. The same way Excel went from a curiosity in the early 2000s to a non-negotiable skill on every job advert, AI is becoming the baseline. The businesses that map it deliberately pull ahead. The ones that wait, or guess, get left behind.

The audit exists to tell you, for Milton-Lloyd specifically: what’s worth implementing, what’s noise, what will quietly save you money — and where AI is actively dangerous and should be kept out.

Three layers. You can’t skip one.

You don’t bolt AI onto a business. You build up to it. Most failed AI projects skip straight to the top layer before the two underneath are ready.

1
Foundation

Is your data AI-ready?

We look at where your data lives and what state it’s in — product catalogues, customer records, seller accounts, order history across every channel. If it’s scattered, old, or unstructured, AI can’t use it. So we optimise the foundation first.

2
Systems

Can your tools even connect?

We audit every system you run — your store platform, Amazon, the seller portal, your CRM, accounting and stock. Some software is so old it has no way to connect to anything else (no API). Where that’s the case, we tell you plainly: this needs to change before AI is possible — and whether that means switching or building custom.

3
Intelligence

Where AI actually goes in

Only once the foundation and systems are sound does AI get implemented — automations, AI co-workers, custom builds. This is the layer everyone wants to start at. It’s also the one that fails when the two below it aren’t ready.

This is the whole point of doing the audit first. It’s a long-term map, not a quick patch. You analyse the entire business from point A (a customer or order arrives) to point B (it leaves) — and you plan in years, not weeks. That’s what makes it a partner relationship, not a one-off tool purchase.

A live platform, not a 20-page PDF

You don’t wait three weeks and get handed a document. You get a login on day one and watch your audit build, department by department.

01
Department-by-department interviews
We talk to the people who actually run each part of the business — production, fulfilment, the seller-network team, customer service, finance — because they know their processes best. The number of calls depends on how many departments you have.
02
A live audit portal
You get login access from the start. After every department call, the platform updates within a day or two — so you see progress live instead of waiting for a final report. Everything exports to PDF if you want it.
03
A full process map + pain points
Every manual step, every tool, every bottleneck — captured verbatim from your team, organised by department and priority (high / medium / low), with the hidden cost of each one estimated.
04
A priority matrix — impact vs. effort
The deliverable that pays for itself: a clear map of what’s high-impact and easy to build. Even if you never work with us again, you can take the quick wins and act on them immediately.
05
Risk mapping
Not just where AI helps — where it’s dangerous. AI isn’t perfect and has real risks. Knowing where not to use it is as valuable as knowing where to.
06
Staff education built in
The goal isn’t to replace people — it’s to make them more efficient. We teach your team how to use AI so they spend less time on manual work and more on the things that actually matter.

Full money-back. No questions asked.

The audit carries all the risk, not you.

Risk reversal

If we don’t find enough waste to justify the audit, you pay nothing.

We refund you in full — and you keep everything we produced. Burn it, or take it to another agency and have them do the work. There is zero obligation to do any further work with us. We’ve delivered 15+ audits and are still to issue a single refund.

Where we’d expect to find it

A fragrance manufacturer selling direct, on Amazon, and through a global seller network has more moving parts than almost any business its size. These are the areas the audit would dig into — the exact numbers come from interviewing your team.

Multi-channel order & stock reconciliation

Own stores (UK / US / Rest of World), Amazon, and the seller network all pulling from the same inventory. Where channels don’t talk to each other, stock counts drift, orders get manually re-keyed, and someone reconciles it all by hand.

The seller network at scale

“Free accounts, set up in minutes” means a high volume of sellers to onboard, support, route orders for, and pay. At scale, manual onboarding and support quietly become one of the biggest time sinks in the business.

Manufacturing & demand planning

Blending and bottling across a large fragrance range means forecasting demand from messy, multi-channel sales data. Better signal here is the difference between stockouts on a best-seller and cash tied up in slow lines.

Marketing & conversion

Sample “pick & mix”, the fragrance-finder quiz, refer-a-friend codes, regional newsletters. Are sample buyers being followed up to a full-size sale? Is referral revenue tracked? This is usually where recoverable revenue hides.

Customer service across regions

Multiple regions and currencies means repetitive enquiries in volume — order status, returns, scent recommendations. A large share is automatable without losing the personal touch.

The systems underneath

Before any of the above can be automated, we check whether your store, marketplace, seller portal and back-office systems can actually connect. Old platforms with no API are the silent blocker on every AI project.

Closest matches to your business

Real audits, anonymised under NDA. We’ve not yet audited a perfume house — but we’ve audited the three things Milton-Lloyd is: a product company with a distributor network, a high-volume customer operation, and a manufacturer running a disconnected stack.

A B2B product distributor

Reps managing a network of accounts

A products business selling into a large network of accounts through field reps — the closest parallel to your seller network. Reps were spending the bulk of their week servicing tiny accounts instead of growing the base, with no visibility on activity or churn.

$104,000/yr waste identified
Largest single item: reps spending 32 hrs/week visiting accounts worth $5–7K/yr each instead of prospecting.
A high-volume consumer operation

Onboarding & revenue leakage at scale

A business onboarding hundreds of new customers a year on a fully manual stack — like a flood of free seller sign-ups with no system underneath. Revenue was leaking because a large share were never properly invoiced or followed up.

$252,760/yr waste identified
One automation recovered $47,923/yr of missed revenue in a half-week build — before the rest of the stack was touched.
A manufacturer

Systems that don’t talk to each other

A manufacturer paying for capable, expensive software — that wasn’t connected. Projects and orders were re-created by hand across three systems, reconciliation ran in spreadsheets, and invoices got double-paid because nothing cross-checked.

$67,600/yr waste identified
“We don’t have any connectivity… we’re not moving any data yet. That’s our problem.”

What we find, business after business

The pattern holds across industries: there is always significant, quantifiable waste hiding in manual process. Figures below are real, anonymised audit results.

15+
Audits delivered
$67K–$957K
Annual waste found per business
0
Refunds ever issued
Business typeQuantified waste / yrTop problem found
Real estate group$957,6002,987 appraisal leads + 1,100 sales opportunities never followed up
Youth sports operator$378,0003 hrs admin per signup; members lost each term to forgotten re-enrolment
High-volume placement agency$252,760Hundreds joining/yr, only a fraction ever invoiced — revenue leaking
Immigration firm$208,0008 staff running operations on Excel and notepads
B2B product distributor$104,000Field reps servicing low-value accounts instead of growing the base
NDIS provider$120,000Ops manager building tomorrow’s schedule by hand, 3–5 hrs/day, 7 days/week
Marketing agency$92,560Founder bottleneck — 6–8 week manual onboarding per client
Manufacturer$67,600Core systems with zero connectivity; invoices double-paid

Client names withheld under NDA. Tailored, named case studies for the fragrance and consumer-goods space are being recorded now and will follow.

Why we don’t quote a build on day one

Anyone who promises you a specific automation before understanding your business is guessing.

You can’t responsibly say “we’ll build you X” before mapping how the business actually runs. That’s the entire reason the audit comes first — it tells us what’s genuinely high-impact and easy to build, instead of selling you something that bites you later.

After the audit, the natural next step is a quick win: something high-impact and easy to build that proves the value fast. Where that leads — more automations, system changes, a custom platform — is entirely your call. The audit is deliberately lean and clean: you own the full picture of your business, and you decide what happens next. There is no obligation to build anything with us at all.

Map the business. Then decide.

Ready to see where AI belongs in Milton-Lloyd?

The next step is a short discovery call to scope the audit around your departments and channels. No commitment — just a conversation about what the audit would cover and what it would cost.

01 Discovery call
02 Scope & agree the audit
03 Department interviews
04 Live portal builds out
05 Priority matrix & quick wins