MilessianMILESSIAN Request a demo
MILESSIAN/ˈmi.lɛsː.i.ɑn/
Marketing & sales decision intelligence

You don’t lack data. You lack decisions you can trust.

Milessian is decision intelligence for marketing and sales. It turns the data you already have into verified evidence and a decision you can act on, next to the systems you already run. Down to a single location and week.

Built for data-rich organizations. Milessian does the reading. You make the call.
See it before the report does.Sustained changes surface while they are still local.
Down to location and week.Region, location, catchment, category, segment.
Campaigns from evidence.Timing, audience, offer and channel checked first.
Effects measured where the data allows.Method fixed before launch. If an effect cannot be estimated, Milessian says so.
01
The problem

The opportunities are already in your data. The way you work keeps them out of sight.

Your company holds years of sales, customer and campaign history, and the world around it adds new signals every day. But the data sits in separate systems and separate teams, and turning it into a decision still happens by hand.

I.

Teams don’t share one picture.

Marketing, sales, category and CRM each work from their own system and their own numbers. Every hand-off between departments costs time, and context is lost at each one.

II.

The work lives in spreadsheets.

Exports with thousands of rows and endless columns, rebuilt for every briefing. There is rarely time to look beyond the next deadline.

III.

Plans start from habit, not evidence.

Last year’s calendar, one platform’s dashboard, a gut feeling. Marketing and sales strategy and campaign planning are rarely built on the data the company already owns.

The chain today

  • 01Data sits with separate teams: CRM, marketing, sales, category.
  • 02Exports of thousands of rows, studied now and then, if at all.
  • 03An analyst builds a one-off deck for one question.
  • 04A decision is made and a brief goes to the agency.
  • 05Months later, a spreadsheet tries to say whether it worked.

Meanwhile, the opportunities were in the data from the start: in what your customers do, in what your history shows, and in what is changing around you.

02 · The answer

Milessian reads them together.

Internal history and external signals, each moving on its own, read as one picture. What comes out is not another report, but opportunities you can plan a campaign on.

  • Sales history
  • Loyalty-card behavior
  • Basket patterns
  • Campaign history
  • Stock & stores
  • Web & app
  • Weather
  • Search interest
  • Competitors
  • Calendar & events
  • Economy & prices
  • Local & demographics

Market opportunities

What is changing around you

Action Plan · who
DIY households in 7 cold-front catchments
With what
Heating campaign: local ads and email, 10-day window
Why
Cold front, rising search interest, stock available
Evidence Strong · 3 independent signals

Customer opportunities

How loyal customers behave

Action Plan · who
Loyalty members who bought tiles but not grout
With what
Personal offer for grout and sealant, in the app and by email
Why
Basket pattern: the project is under way, one item missing
Evidence Strong · 3 independent signals

Forecast opportunities

What is likely to come next

Action Plan · who
Garden buyers in 3 regions
With what
Early spring push, two weeks before the usual peak
Why
In past seasons this signal led the peak by 2–3 weeks
Evidence Strong · 3 independent signals
People decide → Results measured → Milessian learns

Dashboards show what happened. Milessian shows what to do.

Action Plan

Every opportunity comes with a plan: who to reach, with what, and why.

Evidence

Every recommendation carries the evidence behind it, so you can check it before you decide.

A system that learns

Decisions and measured results flow back, so the next plan starts smarter.

03Meet Milessian

What Milessian does with them

It doesn’t stop at reading. It interprets, proves and plans.

A signal on its own is not an answer. For every opportunity it finds, Milessian works out where it is, why it is there, how much evidence stands behind it and what to do next. Down to a single location and week.

1

It interprets

Every signal is compared with its own normal pattern, location by location and week by week. A lasting change is separated from a one-off spike.

2

It proves

Every opportunity carries its Evidence: which signals agree, how independent they are, and what is still unknown and why. In words, never as an invented percentage.

3

It plans

Every opportunity comes with an Action Plan: who to reach, with what, where and when, and how the result will be measured.

OpportunityIllustrative example

Local heating campaign in 7 catchment areas

Retail example · 10-day window · best start in 6 days · where demand is already rising, before it reaches the national figures

Where, locally

Category sales against the usual level for these weeks

  • Catchment 1+24%
  • Catchment 2+21%
  • Catchment 3+19%
  • Catchment 4+17%
  • Catchment 5+15%
  • Catchment 6+12%
  • Catchment 7+9%

The other 16 locations are still within their normal range.

Evidence
  • Category sales above their usual level for three weeks, in 7 of 23 locations
  • Cold front forecast for the northwest within five days
  • Search interest for heating products rising in three counties
  • Stock available in 5 of the 7 stores
  • In the last three seasons the same pattern led the peak by two to three weeks

Evidence Strength: strong. Several independent signals agree.

Not yet known: competitor promotions in two counties. There is no public price data for them yet. Milessian checks again every week and will show it here as soon as it can say.

Action Plan
Who
Repeat buyers of heating products in the 7 catchments, plus households searching nearby
With what
Local ads and email with a heater bundle; in-store display in the 5 stores that have stock
Where and when
7 catchment areas, starting in 6 days, for 10 days
Why
Demand is rising here first, the cold arrives here first, and stock can serve it
How we will know
Targeted areas against comparable areas, with the method fixed before launch
Your team decides. Approve, adjust or decline.The result is measured, and Milessian learns from the decision and the outcome.

Not a report to interpret. A plan to approve.

From complexity to confident decisions.

What Milessian does

·

From signal to opportunity

Three ways to find the opportunity. One loop to learn from it.

Every opportunity has one primary reason to exist. Evidence from the other angles supports it or challenges it, and the same fact is never counted twice.

NOW

What is changing around you now?

Demand, seasonality, competitors and external signals against your own normal pattern.

CUSTOMERS

Where is value waiting in customers you already have?

Purchase cycles, started projects, segments and the next likely need.

NEXT

What is likely to come next?

Early signs that preceded demand changes in your own history, only where that history is strong enough.

LEARNThe shared loop: every decision is measured against a method set in advance; reviewed learning informs the next one.

WHAT WE SEE

Interior-paint demand above its seasonal norm for three weeks — in 7 locations of one region, nowhere else.

WHY IT MATTERS

The window is local and short. Acting now differs from reacting when it reaches the national report.

WHAT WE COULD DO

A local campaign in those 7 catchment areas for the rest of the season — stock checked first.

WHAT WE DON’T KNOW ?

Whether stock coverages two locations. Not estimable until stock data is connected.

HOW WE’D MEASURE

Targeted locations against comparable ones, over a window agreed before launch.

Locations above norm7 of 23
Supportingsearch interest · customer projects
Consecutive weeks3
Challengingnone found
Other regionswithin usual range
Sourcefact ID on every number
Stock, 2 locationsnot estimable ?
Compared withown seasonal pattern
TypeMarket (primary)
Levelcategory × location × week
Numberscalculated in code
Opportunity whensustained · beyond usual · material
Sensitivityset per organization
·

Decision intelligence in action

See a decision being made.

One opportunity, from the first signal to an approved plan.

Swipe to see the full screen →
Today view: three decisions for this week, each with estimated extra sales, a deadline and the strength of the evidence
Step 1 · Today

Three decisions for this week, ranked by value. Each card shows the estimated extra sales, the deadline and how strong the evidence is.

Why it matters: the team knows what to act on today, and what can wait.
Swipe to see the full screen →
Opportunity detail: interior paint demand, with the estimated gain and the cost of waiting
Step 2 · Decide

The full picture of one decision: what is happening, why we believe it, the options with their estimated value, and the cost of each day of delay.

Why it matters: the approver sees the reasoning and the numbers before anything is launched.
SignalSales change in the stores is detected and checked against a normal range.
DecisionOptions with estimated value, deadline and evidence, approved by a person.
MeasurementThe result is tracked against a control group, so the next decision starts from facts.
See this for your decision →

A spike is not an opportunity. A lasting change might be.

How Milessian reads signals

·

Under the hood

What Milessian watches, inside and outside. Read together.

Milessian connects what happens in your business with what happens around it — and compares both with your own normal pattern, location by location, week by week.

Internal signals · your business

Sales by location, category, weekdeviation from the seasonal norm
int
Stock and availabilitycan the opportunity be served?
int
Customers and basketspurchase cycles, projects, next likely need
int
Campaigns and promotionswhat ran, when, where — and what it changed
int
Web and app behaviorvisits, carts, interest before purchase
int
Prices and marginswhere available and authorized
int

External signals · the world around you

Weather and forecastscold fronts, heat, rain — by region
ext
Search interestwhat people start looking for, and where
ext
Competitors nearbyopenings, closures, promotions in the catchment
ext
Seasons, holidays and local eventsschool breaks, holidays, regional dates
ext
Market and economic contextcategory trends, prices, consumer climate
ext
Your own contextnotes and plans your team adds — kept separate from facts
ctx
Fig. 1 Ten signal families as continuous streams. Most of the movement is noise; a few currents converge into something worth a decision.
All ten signal families

A spike is not an opportunity.

One unusual week is noise. Milessian looks for changes that last, exceed the usual variation and matter for the business, and it says why.

Fig. 2 Weekly demand index for one category in one location, against its usual seasonal range.

Hyperlocal, without false precision.

Always the finest level the evidence supports, and never finer.

Fig. 3 From country to a single catchment area and category-week.

Signals are used where available and authorized for your organization. Which sources apply is agreed before any pilot.

Why you can rely on it

Every number has an address.Open any figure to see where it comes from and what it depends on.
Unknown stays unknown.When the data cannot support an answer, Milessian says why and what is missing.
People decide.A recommendation is not an approval. Your team makes the call.
See this for your decision →
04Your role

Decision intelligence by role

The same evidence, answering your question.

Pick your role. See the question Milessian answers for you, what you would see, and what you are measured on.

Leadership
Marketing
Sales & trade
Category & retail
Data & analytics
See this for your decision →
·

Retail, FMCG and other industries

Your industry. Your signals. The same discipline.

The signals change from industry to industry. The logic does not: connect them, find what deserves a decision, plan it, measure it.

Fig. 4 — Signals flowing into one opportunity.

Retail is where Milessian is demonstrated today. If your industry is not listed, “Another industry” shows a neutral example. Industry-specific data and definitions are scoped together before any pilot.

Whatever the industry, the first step is the same: one decision that matters. See how a pilot starts →

05

How a decision intelligence pilot works

Start with one decision that matters.

No big-bang integration. A focused pilot on your data, one business question, and an honest view of what the evidence can and cannot support.

1

Scope the question

One business area, the decisions you face, the signals that matter, how success will be measured.

2

Connect a data slice

An agreed extract of your data, quality-checked on arrival. Gaps are reported, not filled.

3

Decide and measure

Opportunities and plans on your data; your team decides; results measured as agreed.

TYPICALLY NEEDED

  • Sales by location, product and week
  • Campaign and promotion history
  • Stock or availability, where relevant
  • Customer data — only if authorized

YOUR DATA

  • Kept separate from every other organization’s
  • Processed under a written agreement
  • Used only for the agreed purpose
  • Details in the pilot agreement
·Answers

Decision intelligence FAQ

Short answers to fair questions.

What leaders usually ask before a first conversation. Plain answers, no promises we cannot measure.

What is Milessian?

Milessian is a decision intelligence platform for marketing and sales teams. It reads internal and external signals together. It shows which changes deserve a decision, and where. People decide; Milessian prepares the evidence.

Do we need a data science team to use Milessian?

No. Milessian is made for marketing, sales and category teams: it connects the data, runs the analysis automatically and explains each result in plain language. If you have a data team, they keep oversight, because every figure is traceable to its source.

What is decision intelligence?

Decision intelligence connects data, analytics and a defined decision. Instead of another report, it shows which change deserves action, what the options are, and how the result will be measured. In Milessian, every number is calculated by code and traceable to its source; people make the decision.

How does Milessian measure the impact of a campaign?

Before launch, the method is fixed: which locations or customers are targeted, and which comparable ones are not. Afterwards Milessian compares the two, reports the difference with its range, and says when the effect cannot be estimated. There is no revenue guarantee, only an agreed way to measure.

How is decision intelligence different from a BI dashboard?

A dashboard shows what happened. Milessian compares each signal with its own normal pattern, by location and week. It flags only lasting changes that matter for the business. Each one comes with evidence, options and a way to measure the result.

Which signals does Milessian use?

Your own data first: sales, stock, customers and campaigns. Then the context around you: weather, competition, calendar, local economy and search interest. Which sources apply is agreed before any pilot.

Does Milessian change budgets or campaigns automatically?

No. Milessian prepares plans and scenarios. Your team approves them. Execution stays in your own tools.

How does Milessian keep AI from inventing numbers?

Numbers are calculated by code, not written by a language model. Every figure can be traced to its source. Statements are checked against the facts behind them. When the data cannot support an answer, Milessian says so and explains why.

Is Milessian only for physical stores?

No. Chains with physical stores gain the most from catchment-level signals: stock, space and local campaigns. Online stores use the same regional signals to target online spend and to plan stock and delivery by region. Channel effects are shown only where spend history allows.

What data is needed for a pilot?

Usually sales by location, product and week, plus campaign history. Stock and customer data help where they are available and authorized. Gaps are reported, not filled.

How is our data protected?

Each organization’s data is kept separate from every other. It is processed under a written agreement and only for the agreed purpose. The details are set out in the pilot agreement.

Which industries is Milessian for?

Retail is where Milessian is demonstrated today. The same logic applies wherever many local signals shape demand: automotive, banking, consumer health and manufacturing.

What does “not estimable” mean?

It means the data is not enough for a reliable figure. Milessian shows what is missing and what happens next. Missing data is never treated as zero.

The name

Deciding from evidence is not a new idea — just newly within reach.

Thales read the signs and foresaw a bumper olive harvest, so he rented the presses before anyone else could.

MILESSIAN/ˈmi.lɛsː.i.ɑn/with a long, held s
Make the next one.

Read the full story →

In short

One decision that matters. Not another report.

It reads

Your history and the world around you, read together, down to a single location and week.

It proves and plans

Every opportunity comes with its Evidence, what is still unknown, and an Action Plan.

You decide

Your team approves. The result is measured, and Milessian learns from it.

Request a demoA 30-minute conversation about one decision you face. A personal reply within one business day.

06

Request a demo

Make your next decision with a clearer view.

  • Your business context and decision questions
  • The signals you have — and what they can support
  • One decision situation, from signal to plan
  • The product, shown on your decision

“A campaign is a consequence, not a starting point. Marketing belongs at the core of the business decision, not as cosmetics on a decision already made.”

Tamás Fodor, founder · your first conversation is with him
Tell us where to start
Please use your company address. Personal mailboxes are not accepted.

Off-beat Communications Kft. processes your details only to answer this request and arrange a demonstration (GDPR Art. 6(1)(b) and (f)). Your message goes to our team, and we reply from our company address. Details are kept for 12 months after our last contact, then deleted. You can access, correct or erase them at any time — see the privacy notice.

We reply from our company address, usually within one business day.