Skip to main content
Allways

Solutions by sector

The criterion is not the industry: it is the shape of the problem. These are the sectors where we know how to spot it.

Twelve shapes of problem we know how to solve. If your sector is not on the list, the shape of the problem probably is. Tell us the case and we will say on the first call whether we recognise it.

The sectors where we recognise this shape of problem

Retail and commerce

The stock says one thing in the shop and another on the website, and the customer finds out after paying.

One inventory for shop and website, a catalogue whose prices and promotions publish themselves, and a checkout step that does not go down on campaign day.

What we measure

  • Orders cancelled because the stock was not really there
  • Minutes between the sale at the counter and the website update
  • Orders completed per hour at the campaign peak

What shapes the design

  • Consistency across channels
  • Campaign peaks
  • Integration with the point of sale

How far this can go

  1. Short cycle: the stock in the shop and the stock on the site become the same number.
  2. With integration: the till, the catalogue and the checkout talk to each other without anyone exporting anything.
  3. With our own research: restocking forecasts per product and shop, when there is enough sales history; when there is not, stock-out alerts first.

Real estate

The inquiry arrives on Saturday night and someone answers it on Monday at midday. By then they have already visited two other properties.

A portal that serves heavy photos without blocking the page, a property page search engines understand, inquiries routed to the agent who is free, and a booking signed online.

What we measure

  • Minutes between the inquiry and the first reply
  • Inquiries per published property
  • Load time of the property page on a phone using mobile data

What shapes the design

  • Competition in local search
  • Heavy media
  • Traceability of inquiries

How far this can go

  1. Short cycle: the inquiry reaches whichever agent is free, and is logged with the time it arrived.
  2. With integration: the portal, the property record and the online signature share one source.
  3. With our own research: price estimates from local data, when there are enough comparable transactions; when there are not, comparables chosen by hand first.

Hospitality and tourism

Bookings come in through five platforms that charge commission, and availability is synced by hand.

A direct booking engine, one calendar connected to the channels, deposit payments, and a cancellation policy applied by the system and not by a person.

What we measure

  • Direct bookings over total bookings
  • Overbooking cases per month
  • Load time of the payment step

What shapes the design

  • Overbooking
  • Seasonality
  • Payments, deposits and cancellations

How far this can go

  1. Short cycle: direct bookings stop depending on someone watching five dashboards.
  2. With integration: one calendar pushes availability to the channels and takes their bookings back.
  3. With our own research: pricing by season and occupancy, when there are two full cycles of history; when there are not, season rules you can revise by hand first.

Logistics and distribution

Routes are put together in a spreadsheet, and the status of a shipment is found out by calling the driver.

Route building with the real constraints of the fleet, a driver app that works with no signal and syncs when it comes back, and public tracking for whoever is waiting for the parcel.

What we measure

  • Deliveries resolved on the first attempt
  • Minutes of planning per route
  • Status checks that end in a phone call

What shapes the design

  • Intermittent connectivity
  • Location data
  • Integration with carriers

How far this can go

  1. Short cycle: the status of a delivery stops being a phone call to the driver.
  2. With integration: the route, the delivery note and the external carrier share one status.
  3. With our own research: routing built around the real constraints of your fleet, when there is history of time per stop; when there is not, zone and window rules first.

Training and private education

The platform holds up all year and goes down the week of enrolment.

Enrolment sized for the peak, a campus with materials and submissions, assisted grading, and a progress panel per student.

What we measure

  • Enrolments completed over enrolments started, at peak hour
  • Server errors during the enrolment window
  • Students who drop out before the first submission

What shapes the design

  • Seasonal load peaks
  • Accessibility as a requirement, not an extra
  • Data on under-age students

How far this can go

  1. Short cycle: enrolment holds up through the peak hour without falling over.
  2. With integration: the campus, the payment gateway and the student record stop duplicating the sign-up.
  3. With our own research: early warning of drop-out, when there are enough past cohorts with known outcomes; when there are not, alerts on missed submissions first.

Professional services

You bill by the hour, and the hours go into gathering documents, asking for them again, and working out which one was the latest version.

A portal where the client uploads their documents, data extraction from whatever arrives, version control, and a case file with a history of who saw what.

What we measure

  • Billable hours over hours worked
  • Rounds of document requests per case file
  • Days between opening the case file and the first delivery

What shapes the design

  • Confidentiality of the documents
  • Retention periods
  • Access logging

How far this can go

  1. Short cycle: the client uploads their paperwork once, and it keeps its version and its date.
  2. With integration: the case file, the invoicing and the time tracking read the same record.
  3. With our own research: data extraction from the documents that arrive, when there is volume and the formats repeat; when there is not, templates with assisted checking first.

Industry and production

The machine knows why it stopped, but that stays on its own screen and the production report is filled in by hand.

Machine data capture, a production report that fills itself in, stoppage alerts with their cause, and a dashboard that compares shifts and lines.

What we measure

  • Minutes of unplanned downtime per shift
  • Time between a stoppage and the alert reaching whoever can fix it
  • Production reports closed without going through paper

What shapes the design

  • Machines of different ages and protocols
  • Plant floor without stable coverage
  • The plant network is not opened to the internet

How far this can go

  1. Short cycle: the production report leaves paper behind and is visible within the same shift.
  2. With integration: machine data arrives on its own, over OPC UA or Modbus, if the manufacturer exposes it.
  3. With our own research: failure prediction, when there is enough history of labelled faults; when there is not, we instrument and measure first.

Site and field work

The crew finishes on site and the report arrives three days later, on paper, with the photos on someone's phone.

A field app that works with no signal and syncs on the way back, photo and client signature on site, and crew planning by area.

What we measure

  • Hours between finishing the job and closing the report
  • Jobs that need a second visit
  • Reports that arrive with no photo or no signature

What shapes the design

  • Hours with no coverage
  • Devices used with gloves and in full sun
  • A record of who did what, and when

How far this can go

  1. Short cycle: the report is closed on site, with photo and signature, before the crew gets back.
  2. With integration: the report reaches invoicing and maintenance without anyone typing it twice.
  3. With our own research: crew planning around the real constraints of the operation, when there is history of time per job type; when there is not, rules by area first.

Traceability and compliance

When someone asks what happened to a batch, it has to be pieced together from emails, spreadsheets and whoever was there.

An event log that cannot be edited after the fact, a file per batch or per operation, and history exported in the format the auditor asks for.

What we measure

  • Hours to reconstruct the history of one operation
  • Events logged without manual entry
  • Audit requests answered without asking another team for data

What shapes the design

  • Retention periods that differ by type of data
  • Access logging
  • The format of the report is set by whoever audits

How far this can go

  1. Short cycle: the events that live in email today move to a log with a date and an author.
  2. With integration: the history is built from what your systems already emit, without duplicating the record.
  3. With our own research: anomaly detection over the history, when there is volume and labelled cases; when there is not, rules you can explain one by one first.

Case files and procedures

The file passes through four people and nobody knows which one has it: whoever filed it calls, and someone goes looking.

A case file with a status the filer can see, routing and a deadline per stage, automatic notices, and a history of who saw and who decided what.

What we measure

  • Days between opening a file and resolving it
  • Files held up longer than their own deadline allows
  • Status enquiries that end in a phone call

What shapes the design

  • Deadlines set by regulation, not by the design
  • Third-party personal data inside the file
  • Whoever files it may have no account and no app

How far this can go

  1. Short cycle: the status of a file stops being a phone call and becomes visible from outside.
  2. With integration: the file feeds itself from the registry and the document manager you already use, instead of asking twice.
  3. With our own research: automatic data extraction from the paperwork that arrives, when the volume justifies training it; when it does not, templates and assisted checking first.

Associations and community

The membership roll lives in a spreadsheet and nobody knows who stopped paying until the money is missing.

Joining and leaving online, dues collected on their own with a receipt, arrears notices, and a roll that does not live on one person's computer.

What we measure

  • Dues collected without manual entry
  • Days between a missed payment and the first notice
  • Sign-ups completed against sign-ups started

What shapes the design

  • Each member's consent to be contacted
  • Dues with bands, discounts and family rates
  • A committee that rotates and has to get in without depending on anyone

How far this can go

  1. Short cycle: the roll leaves the spreadsheet and the dues collect themselves.
  2. With integration: the accounts and the payment gateway receive the record once, not three times.
  3. With our own research: early warning that a member is about to leave, when there is enough history of joining and leaving; when there is not, rules on arrears and attendance first.

Media and content

The catalogue grows, every contract expires on a different date, and to know whether something can be published you open the contract.

A catalogue with rights as data —window, territory and format—, publishing that respects that window on its own, and search that understands the archive.

What we measure

  • Items published outside their rights window
  • Minutes to find out whether an item can be published
  • Archive items that actually come up when you search for them

What shapes the design

  • Rights that differ by territory and by format
  • An old archive that is poorly described
  • Traffic peaks on release day

How far this can go

  1. Short cycle: rights stop living inside the contract and become a field the system can check.
  2. With integration: the catalogue syncs with the player and with whoever distributes, with no spreadsheet in between.
  3. With our own research: automatic description of the old archive, when there is enough material to evaluate it; when there is not, assisted cataloguing first.

How we decide whether your problem has this shape

The sectors above share four traits. If you recognise three, we can probably help you, even if your industry is not listed.

  • The same data lives in more than one place, and the places do not agree.
  • There is one step of the process only one person knows how to do, and that person takes holidays.
  • The system holds up all year and goes down on the week that matters most.
  • Nobody can reconstruct what happened last week without asking someone.

Is your sector here? See the full list: 88 divisions across 21 sections.

The criterion is not your industry: it is the shape of the problem. That is why the list is complete, and why each division links to the shape that recognises it.

There is a shape for 84/88 divisions today. The other 4 stay on the list on purpose: having no card does not mean we do not work there, it means there is not yet a shape that sums them up. Tell us the case and we will say whether we recognise it.

AAgriculture, forestry and fishing

BMining and quarrying

CManufacturing

DElectricity, gas, steam and air conditioning supply

EWater supply; sewerage, waste management and remediation activities

FConstruction

GWholesale and retail trade; repair of motor vehicles and motorcycles

HTransportation and storage

IAccommodation and food service activities

JInformation and communication

KFinancial and insurance activities

LReal estate activities

MProfessional, scientific and technical activities

NAdministrative and support service activities

OPublic administration and defence; compulsory social security

PEducation

QHuman health and social work activities

RArts, entertainment and recreation

SOther service activities

TActivities of households as employers; undifferentiated goods- and services-producing activities of households for own use

  • 97Activities of households as employers of domestic personnelno card yet
  • 98Undifferentiated goods- and services-producing activities of private households for own useno card yet

UActivities of extraterritorial organizations and bodies

Classification: ISIC Rev.4, from United Nations Statistics Division. The codes and the structure are theirs and can be checked; the labels are our working translation.

Your sector is not here?

That is fine. Tell us what is going on and we will say whether it has the shape we know how to solve. If it does not, we will say so and save you the meeting.

Book 30 minutes