Role
Prices and discounts
Who knows what your largest client really costs after all the discounts, bonuses and retro rebates? If it has to be calculated rather than read off, then pricing decisions are being made without that number.
Intro call
Four questions we would start with
Who approves a departure from the price list?
If the answer depends on the amount, good. If it depends on who happens to pick up the phone, the price list is a suggestion rather than a policy.
How do you know a given discount still has a reason behind it?
Discounts are granted for a circumstance - volume, entering a new segment, compensation for a complaint. The circumstance passes, the discount stays. After three years nobody remembers what it was for.
What do you do when the price list and the system say different things?
Usually what is in the system wins, because that is where the invoice comes from. The price list then becomes an archive document that everyone knows is out of date.
How many people touch one discount decision before the client gets an answer?
The salesperson, their manager, sometimes controlling. Every handover is hours, and in that time the client is talking to the competition.
We take over the calculating, you keep the decision
Assembling the full picture of the price, checking it against the policy and preparing the material for the decision is done by the machine. Whether to grant a departure and on what terms is settled by a person.
- A price is never set automatically. - The system shows what follows from it - the margin after every component, the comparison with other clients, the history of departures. The decision stays with you.
- Every departure has its reason and its date recorded. - This is the one change that makes the biggest difference after a year: a discount stops being permanent, because it has a review date.
- There is one pricing policy. - If a rule exists, the system applies it uniformly and it is visible when someone works around it. If it does not exist, the first summary will show that.
Scope
What can be taken off pricing work
Nineteen example tasks. Each one has a human approval point. Below is a selection, and the scope we fit to your process.
The price list and its versions
- Comparing price list versions and pointing out what changed
- Detecting lines whose price has not changed for a long time
- Checking the price list against purchase cost after a supplier's change
- The list of clients affected by a change and needing notification (the wording of the letter: the specialist)
- The prices of the same item across different clients
Requests for a departure
- Preparing the material for the decision: the margin after the departure, the client's history, similar cases
- Pointing out the approval level required for a given amount (decision: an authorised person)
- Detecting requests that repeat for the same client
- A register of granted departures with the reason and the review date
- A reminder about departures whose date has passed
Margin control
- The margin on a transaction after every component: discount, bonus, retro rebate, delivery cost
- Pointing out transactions below the threshold set for the category
- Clients whose margin falls despite rising turnover
- Detecting lines sold below cost
- Material for a client profitability review (conclusions: the specialist)
Retro rebates and bonuses
- Calculating the bonuses arising from the contracts and pointing out discrepancies
- A reminder about thresholds a client is close to reaching
- Checking the calculation against the wording of the contract
- The cost of bonuses against the turnover they generated
A day's work
Today
- A discount request arrives by email and the margin has to be calculated from scratch
- You look for a client's discount history in old emails
- You find out by chance that a discount granted for a quarter is in its third year
- The margin after bonuses is calculated once a quarter, in a spreadsheet
- When the price list changes you work out by hand who has to be notified
- The question "how much do we really earn on this client" takes a day's work
With AI assistance
- The request arrives with the margin calculated and this client's history of departures
- You can see who was granted a similar discount and with what result
- A departure past its date reports itself
- You have the margin after every component as you go, not quarterly
- The list of clients to notify is produced when the price list changes
- You answer the question about a client's profitability during the conversation
- A consistent price list
the same product does not have four prices without a reason
departures are visible side by side rather than scattered through transactions
- Control over departures
a discount has a review date
granting it records the reason and the date, so continuing it takes a decision rather than forgetfulness
- Time to a decision
the client gets an answer the same day
the material for the decision is produced with the request, not after it
- A record of the reasoning
it is known why this client has this price
the reason is recorded with the decision rather than reconstructed years later
We start with one product category, not with the whole price list.
- Step 1
We assemble the real price
We take a few clients and work out what they really pay after every component. That one table is usually enough for the conversation about pricing policy to stop being theoretical.
- Step 2
A register of departures
We start recording the reason and the date with every new departure. Nothing retrospective - the past cannot be reconstructed and it is not worth the time.
- Step 3
Material for the decision
A discount request arrives with the margin calculated and the history. The approval levels stay as they are - what changes is what someone approves on the basis of.
- Step 4
Reviews
Departures past their date, margin below threshold, bonuses against turnover. This stage only makes sense once the register from step 2 has a few months of history.
Common questions
Questions that come up at the first meeting
Is this a prelude to removing the role?
The opposite - to someone in it finally doing what the title says. Today a pricing specialist mostly recalculates requests and looks for what was agreed last time. When that stops taking a day, the work that there was never time for begins: looking at which discounts eat the margin and where the pricing policy has drifted apart from practice.
Does this fall under the AI Act?
Setting prices for business clients is not a high-risk use and the AI Act does not classify it as one. The limit worth remembering lies elsewhere: automatically differentiating prices towards consumers by their characteristics falls under article 22 of the GDPR (no decisions based solely on automated processing) and under the duty to disclose a personalised price from the Omnibus Directive. We do not do that whatever the regulation. What does follow from the AI Act here is an obligation that applies to every rollout: ensuring AI literacy among the people who use the system.
Still have a question that is not here? Book a call
Next step
How many of your discounts have outlived the reason they were granted for?
We will show you what a pricing decision looks like when the margin after every component can be read off rather than calculated.