Home · Solutions · Sales & marketing
Solution · Sales & marketingThe best leads in the group are on its own ramps, and sales is the last to know
The next car offered while the current one is on the ramp
Every car in the workshop diary is scored against rules the directors agree; the ones that qualify reach the right sales adviser in Teams, with history and consent, while the car is still on site.
Executive summary
Your owner base drives through your workshops every month while sales buys leads from portals.
What we deliver is a trigger table and a queue, not a recommendation engine.
Every car in tomorrow's diary is scored before it arrives, so the conversation happens while the owner is in the building rather than in a phone call…
the opportunity and outcome record in the CRM; the vehicle record in the DMS; the Power BI semantic model
Business problem
Owner base
A dealer group sells the car once and earns the rest of its result afterwards: services, parts, tyres, the finance renewal, the used car the vehicle becomes and the next new one. All of it depends on staying in front of the same owner, and the group already has that owner in the building several times a year. Those visits belong to the service department, which runs on different systems and a different diary from the showroom twenty metres away.
The facts that would make the conversation obvious are scattered: the repair order and odometer in the brand's DMS, the contract end date and mileage limit at the finance desk or in a partner's portal, the last offer and the consent record in the group CRM, the car's current value with the used-car desk. No service adviser assembles that between two receptions, and the sales adviser who could use it does not know the car is on site. So the base drives past the showroom to the workshop entrance, and the funnel is filled with bought leads.
How it works today
- PersonThe service adviser opens the repair order, agrees the work and prints the estimate; no contract end date appears on that screen
- WaitingAnything worth a sales conversation waits until the adviser remembers it, which tends to be a quiet afternoon rather than the day the car is in
- PersonOnce or twice a month somebody exports a list from one brand's DMS into Excel and works through it between other tasks
- Risk of errorAn owner with an open complaint, or a car in only for a recall, is approached anyway, because the export carries no such flag
- SystemSales buys portal leads to fill the funnel, and no report sets their cost beside the base already on the ramps
Why the current process costs more than it appears
Nobody planned this work; it accumulated.
- Timing carries the whole value and is the first thing lost. A conversation at the counter, with the car in pieces, costs nothing to start; the same one a week later begins with a phone call that has to justify itself.
- Between the finance file, the DMS and the CRM sits around twenty-five minutes of assembly per car. It belongs to no role, so it happens for the vehicles somebody recalls and for none of the others.
- Approaching the wrong owner has a price nobody counts. An owner told about a new model while arguing over a warranty repair remembers it, and the satisfaction score the importer collects afterwards does not separate the two conversations.
Cost of inaction
Replacing a car is a decision an owner makes once every few years, and the group either takes part in it or hears about it afterwards. The first two rows price desk work and nothing else. The third is an assumption to argue with: no market source supports either figure and both belong to the reader, but leaving the row out understates the case more than including it overstates it. Two hundred and fifty owners is four in a hundred of the qualifying cars a year.
What no row prices is the rules themselves. Without an outcome recorded against every handoff, which cars are worth approaching stays an opinion, settled by whoever speaks last.
A plausible organisation with realistic proportions. The figures are there to be recalculated on your data; they are not a client result.
A dealer group in Poland: four brands, nine showrooms, eight authorised workshops, one DMS per brand, a group CRM, a finance and insurance desk at each site, Microsoft 365 with Teams.
5,600 workshop visits a month across the eight sites. About 9% of them, some 504 cars, meet at least one of the handoff rules the directors have agreed.
The service adviser sees the car and the estimate, contract end dates sit at the finance desk, a list is exported when somebody has time, and sales buys portal leads.
Roughly 25 minutes per car to assemble contract, mileage, history and value into something a sales adviser can act on. Nobody owns those minutes, so the picture exists for a handful of cars.
Robots score every car in the diary against a trigger table the two directors own, apply the suppression and consent rules, and open a handoff task for the named sales adviser in Microsoft Teams while the car is on site.
In the modelled case every car is scored before it arrives, 210 hours a month of assembly work disappear, the handoff reaches an adviser while the owner is in the building, and each rule carries an order and decline rate within a quarter. None of this was measured at a client; it is arithmetic on the assumptions above.
Proposed solution
What we deliver is a trigger table and a queue, not a recommendation engine. A list in Microsoft Lists holds a row per rule: a finance or lease contract ending inside an agreed window, a guarantee ending, mileage running ahead of the contract limit, a repair estimate above an agreed share of the car's value band, a successor model in the group's stock. Each row carries its threshold, the sites it covers and the adviser queue it feeds. A second list holds the suppressions: an open complaint, an insurance repair, a recall-only visit, an owner who has objected, anyone approached by any brand inside the agreed window. Both belong to the two directors.
Robots do arithmetic and carry the result. Overnight, and again as each repair order is opened, a robot pulls the diary, the order, the contract end date, the mileage limit, the odometer trend and the consent state from the systems that hold them. Where a rule fires and no suppression applies, a UiPath Action Center task reaches the named sales adviser in Microsoft Teams with the car, the owner, the history and the channels consent permits. Nothing about the repair is judged by a robot: the technician and the service adviser decide what the car needs, and the sales adviser decides whether there is a conversation to have. The answer, including "wrong moment", is written back against the vehicle.
UiPath Orchestrator queues, triggers, credential store and run log; UiPath Integration Service connectors for Microsoft Teams and Microsoft OneDrive & SharePoint; UiPath Action Center actionable notifications completed in Microsoft Teams; Microsoft Lists versioning; Power BI as a Teams tab
The trigger and suppression tables, the scoring run per brand, the handoff task and its outcome form, the write-back, the adviser queues, the daily site list and the Power BI rule model
Each brand's DMS, the finance register and the CRM through the vendor's API where one is offered, otherwise UI automation under the robot's own account; the valuation source the group already subscribes to
How the automated process works
- AutomationEach evening a robot reads tomorrow's diary from every DMS and queues one item per vehicle in Orchestrator; a repair order opened during the day is queued as it happens
- AutomationThe robot assembles the picture per vehicle: contract end date and mileage limit from the finance register, odometer trend and history from the DMS, last contact and open cases from the CRM, value band from the valuation source
- AutomationThe trigger table is applied exactly as written; where a rule fires the suppression list is checked next, so a recall-only visit, an insurance repair or an open complaint is dropped before any task exists
- AutomationThe consent record is read per channel and per legal entity, and the task states which channels the group may use; a car with no channel consent is handed over as a counter conversation
- PersonThe named sales adviser opens an Action Center task in Microsoft Teams showing the car, the owner, the rule that fired and today's work, decides whether to approach, and records the outcome
- AutomationThe outcome is written back to the CRM and the vehicle, the site channel receives the day's handoff list, and the Power BI tab shows order and decline rates by rule and site
Human-in-the-loop model
Automation handles
- Reading every vehicle in every diary and assembling the contract, mileage, history and value picture behind it
- Applying the trigger table as written, then the suppression and consent checks, before any task reaches a person
- The write-back of every outcome, the site's daily handoff list and the order and decline rate per rule
People decide
- What the car needs today: the technician and the service adviser, never a robot
- Whether there is a conversation to have, when and on which channel: the sales adviser, from the task
- Which rules exist, at what threshold and with which suppressions: the two directors, in a table they can change
Before and after
Systems and integrations
We do not add technology to make an architecture look serious. Every element below has a specific job in this process.
Inputs
- workshop diary and repair orders in each brand's DMS
- contract end dates and mileage limits in the finance register
- odometer readings and service history
- consent record and contact history in the CRM
- the trigger and suppression tables in Microsoft Lists
Automation layer
- UiPath Orchestrator
- UiPath Robots
- UiPath Integration Service
- UiPath Action Center
Target systems
- the opportunity and outcome record in the CRM
- the vehicle record in the DMS
- the Power BI semantic model
Human touchpoints: handoff tasks in Microsoft Teams; the daily list in each site channel; the tables in Microsoft Lists; the Power BI rule view
Technologies used
nightly and intra-day runs across four DMSs, one queue item per vehicle, triggers, credential store, run log
Athe handoff task with the vehicle and consent picture, and the outcome form
Asite channel posts and the rule tables read as SharePoint lists
Athe trigger and suppression tables, versioned, owned by the two directors
Aorder and decline rate per rule, brand and site, as a Teams tab
Adiary, repair order, odometer, contract end date, consent record, opportunity and outcome
CIllustrative economic model
What it is worth, with the arithmetic shown.
What the table prices is the noticing, and only the noticing: reading a repair order, digging a contract end date out of the finance file, checking the odometer against the limit and writing a usable note to a named adviser. The conversation, the appraisal and the offer are sales work either way and stay outside the model. Nothing here was measured at a client; 9% of visits meeting a rule, 25 minutes and €25 an hour fully loaded are assumptions to replace with your own.
Run the numbers on your data
An illustrative estimate from your own inputs. It models released capacity; it is not a promise of savings.
Business benefits
- Every car in tomorrow's diary is scored before it arrives, so the conversation happens while the owner is in the building rather than in a phone call a week later
- The modelled 210 hours a month of exporting and cross-checking return to the counter and the showroom floor; what reaches an adviser is a car, a rule and a history
- Approaches the group should not make are stopped by a suppression list, not by whoever happens to know about the complaint
- Every rule earns its place from recorded outcomes, so the sales meeting argues about order and decline rates instead of instincts
The management view
- Handoffs, approaches and outcomes become a monthly figure per site and brand, which makes the group's own base comparable, cost for cost, with the leads it buys
- Sales and aftersales are measured on the same event: the service department's part in a vehicle sale is recorded rather than asserted at bonus time
- A written trail sits behind every approach: which rule fired, what the consent record said, who decided and what the owner answered
Board-level KPIs
Security and governance
The automation holds exactly the rights it needs, and not one more.
- Each robot signs in with its own account per system, allowed to read the diary, the repair order, the contract register and the consent record and to write an opportunity and an outcome, nothing beyond that; thresholds and suppressions are versioned in Microsoft Lists, so the rule behind any approach can be produced as it stood on the day
- Queue items carry a registration number, a rule code and a site; names and contact details are read when the task is created rather than held in the automation layer, the robots run from the EU region of UiPath Automation Cloud, and the lists, tasks and Power BI model stay in the group's tenant
- The consent record is read by the robot and never written by it; which basis and which channel an approach rests on is settled per legal entity with counsel, and an objection recorded anywhere takes that owner off every brand's list on the next run
Why now
The base is unusually rich: 597,400 new passenger cars were registered in Poland in 2025, 8.3% more than in 2024, and 411,000 of them went to company and institutional buyers (PZPM and KPMG, February 2026). A car bought on a contract comes with a date, and that date decides much of who sells the replacement
The replacement decision now carries a powertrain question: 350,500 alternative-drivetrain cars were registered in the same year, 20% more than in 2024 (same source). That is a conversation, and the workshop counter is the cheapest place to start one
Nothing in the design is exotic: Orchestrator triggers and queues, Integration Service connectors, Action Center tasks inside Teams, Power BI as a tab. The group-specific work is the rule table and the connection to each DMS, and doing the same work by hand costs the modelled €5,250 a month
Relevant executive roles
The base the group already owns becomes a measured source of sales conversations, set beside the cost of bought leads
The service department's part in the next sale is recorded rather than argued about, and no owner is approached while their car is in for a recall or a complaint
Advisers receive cars with a reason attached instead of monthly lists, and rules that produce orders are separated from rules that produce declines
Common questions and objections
They do, and that part is worth keeping. What the design removes is the noticing, which today depends on one person remembering one file. The judgement stays with the adviser and now arrives with the contract date and consent state attached.
That is what the suppression list is for, and why every handoff carries a wrong-moment answer. A recall visit, an insurance repair or an open complaint never produces a task, and a rule whose declines pile up is retired rather than defended.
In most groups they are not; they sit with the finance desk or in a partner's portal. The design reads whichever register you keep, and the pilot starts with the rules whose data is already reliable, usually guarantee end and mileage against the limit.
When this is not the right solution
- A single site where service and sales sit ten metres apart and one manager sees every repair order; a short daily conversation costs less than a robot
- No usable record of contract end dates, guarantee terms or odometer readings anywhere in the group, in which case the rules would fire on guesses and the data comes first
A question for the next management meeting
More of our customers stood at a service counter last month than walked into a showroom: how many of them left with a reason to come back to the sales floor, and what was that decision based on?
Implementation approach
What we deliver, and what we need from you to start.
We deliver
- Two working sessions with the two directors to write the first trigger table: rules, thresholds, windows, adviser queues and suppressions
- A dry run over three months of historical diaries showing which cars each rule would have flagged and what those owners did next
- The scoring run per brand: DMS and finance-register reading, the consent check, the queue and the Action Center task in Teams
- The outcome form, the write-back to the CRM and the vehicle record, and the Power BI rule model
- A pilot at one site with two rules, then the remaining rules, brands and sites, with hypercare
We need from you
- Three months of diary and repair-order history for one brand, with odometer readings and contract end dates as you hold them
- A sales director and an aftersales director to own the trigger table, plus a site director and an adviser for the pilot
- Robot accounts for the DMS, the finance register and the CRM, and a Teams team per site
Stages
Discovery
Diary, repair orders, contract data and the consent record as they actually exist, at one site
Design
Trigger table, thresholds, suppressions, adviser queues, outcome form, security model
Build
Scoring run per brand, consent check, Action Center tasks in Teams, write-back, Power BI model
Pilot and rollout
One site and two rules for a quarter, then the remaining rules, brands and sites
Quick win. Effort depends on how many DMSs hold the diary, whether contract end dates exist in a readable register at all, and how many suppressions the consent structure needs.
The customer whose lease ends in March is on ramp four this morning.
Send us one month of workshop diary lines from a single brand, with odometer readings and contract end dates as you hold them. We come back with the cars the rules would have flagged and the ones we would have suppressed.
Score last month's workshop diaryThe neighbouring process usually has the same problem
Nine channels, four brands, one receptionist deciding who calls the customer back. The Friday lead waits until Monday.
View solution Customer serviceReminders that bring the car back to your workshopStop learning that a customer has left only when the fourth service fails to appear in the diary.
View solution Sales & marketingA trade-in price in twenty minutes, reconditioning includedThe customer wants a price for the old car today. The used-car manager is in a meeting, and the offer arrives tomorrow.
View solutionIndustries we deliver this in most oftenAutomotive retail