Machine Service Contract: What Has the Biggest Impact on Downtime

Why the contract price does not show downtime risk

The lowest service rate looks attractive only at the moment of purchase. When the machine stops, people no longer count the price of the contract—they count the hours of lost work.

One day of downtime can easily wipe out the discount you got when signing. The workshop loses not only output. Shifts get pushed back, the operator waits, the next batch is delayed, and an urgent order gets moved to the evening or the weekend. If the breakdown happens in the middle of a production run, the losses grow even faster: some blanks may be scrapped after an emergency stop, shipping deadlines shift, and the supervisor has to manually rebuild the entire production plan.

That is why you should not look at the annual contract amount, but at the time it takes to get the machine back to work. If one supplier promises a response within 2 hours, remote diagnostics the same day, and a clear engineer dispatch procedure, while another simply offers a price 15% lower, these are not equivalent offers.

A low price often means the contract has removed the things that affect downtime the most: the first response time, the list of included work without extra charges, out-of-town visits, electronics and CNC diagnostics, and access to common spare parts. In the end, cheap service often covers only a phone call, a basic consultation, and a visit "if possible". Everything that actually gets the machine running again is billed separately.

For CNC machine maintenance in Kazakhstan, the difference is especially noticeable if production is not in Almaty or Astana, but in a region. There, the service response time and engineer dispatch terms affect downtime more than a discount in the contract.

A useful rule of thumb is simple: ask how long it takes from the request to normal spindle or axis operation, not to "request accepted". That is the real risk indicator.

How to write the response time into the contract

The response time in the contract should be counted not from the moment the technician "learned about the problem", but from the moment the service received the request through an agreed channel: email, CRM, messenger, or a phone call with request registration. If that moment is not fixed, the dispute will start at the first shutdown. The client will say they called in the morning, and the service will say the request was only logged in the afternoon.

It is better to split two events right away: the first response and the actual start of work. The first response can be fast and remote. For the workshop, that is not the same as an engineer visiting the site.

Remote support often resolves minor issues in 20–40 minutes: check alarms, parameters, sensors, and basic settings. But if the machine stopped because of an emergency shutdown, you need a separate timeframe for dispatch or the start of diagnostics. A vague phrase like "we respond promptly" is useless here.

What to write instead of vague wording

Words like "as soon as possible", "if possible", and "immediately" are better removed. Before the first breakdown, they sound fine; after that, they mean nothing.

Instead, the contract needs numbers and clear events:

• first remote response within 30 minutes during business hours;
• remote diagnostics within 2 hours after the request is registered;
• emergency spindle or axis shutdown - a separate priority, for example dispatch within 24 hours;
• requests at night, on weekends, and on holidays - under a separate standard stated directly in the contract;
• if repair requires both an engineer and spare parts, the service confirms the action plan within the same timeframe.

Also define the service hours separately. A 2-hour response on Wednesday and a 2-hour response on Sunday are different situations, and the contract should reflect that.

If the supplier also handles sales and commissioning together with service, it is convenient to tie the response time to the machine type and the service region. One standard can apply to Almaty and Shymkent, another to a remote site. That is fairer than promising the same time to everyone.

A good clause usually sounds dry and very specific. That is exactly what later saves a day of downtime, and sometimes an entire shift.

What work should be included in the contract

The most common problem is simple: the contract says "repair and maintenance", but does not say exactly what the service does on site. Then it turns out the engineer came only to "take a look", while root-cause analysis, adjustment, and trial run are billed separately. For a workshop, that is a bad scenario, because the downtime lasts longer than the repair itself.

The contract should include initial diagnostics with explanation: inspection, error reading, checking the mechanics, electrics, and CNC system. If this point is missing, any visit can easily be turned into a consultation.

Also state the root-cause search on site. This is not a formality. One thing is to replace a sensor when the failure is obvious. Another is to catch an intermittent error when the machine sometimes works and sometimes stops. In such cases, the workshop loses hours on finding the cause, not on replacing the part.

What should be listed directly in the text

It is better when the contract names the work clearly and without vague wording. Usually, it is enough to include the following items:

• initial diagnostics and fault localization;
• disassembly, replacement, and reassembly of agreed assemblies;
• parameter adjustment after repair;
• trial run under load or on a test part;
• a brief report with the cause of failure and the list of completed work.

You also need a list of assemblies. Otherwise, the service may say that the mechanical part is included, but the spindle, drive, hydraulics, or lubrication system are not. If the machine runs in two shifts, such gaps become expensive very quickly.

After the repair, the service should not just "restore operation" but bring the machine to normal mode. That is why it is worth adding a clause about adjustment, accuracy checks within acceptable limits, and a trial run. Otherwise, you take on extra risk: the machine starts, but an hour later it stops again or begins producing defects.

Where confusion often appears

Preventive inspection and emergency callout are better kept separate. A scheduled visit usually includes cleaning, wear checks, tightening, lubrication, and replacement recommendations. An emergency callout is about finding the cause of failure and restoring operation. If you mix these formats, a dispute will appear at the first serious failure.

When the supplier handles the full cycle—from equipment selection to service—this list is especially important. The workshop has fewer gray areas, and the service team has fewer reasons to argue that something "was not included in the contract".

What to check for spare parts

A machine usually stops not because of the price of a part, but because the needed item is not in stock or takes too long to arrive. That is why the spare parts section must be specific. The phrase "supply of components" does not help when the CNC lathe is already down.

Separate consumables from expensive assemblies right away. Consumables are replaced regularly: filters, belts, sensors, seals, grease, small relays. Expensive assemblies are bought less often, but they create the longest downtime: spindle, servo drive, boards, motors, pumps, CNC components. If the contract does not divide these groups, lead times and delivery rules almost always become a dispute.

The contract should state:

• which items the service keeps in stock at all times;
• which parts are supplied only to order;
• what lead time is promised for common parts;
• who confirms availability in writing;
• who is responsible if the part does not fit.

Ask for lead times not in general terms, but for several common items. A sensor or belt has one lead time; a servo drive or board has another. The phrase "within a reasonable time" is worth nothing here. You need numbers: 2 days, 10 days, 30 days.

Also clarify where the warehouse is located. For a company in Kazakhstan, this changes the real picture a lot: the part may be in a local warehouse, with a partner in another country, or only at the manufacturer. These are three very different scenarios in terms of timing and costs. It also helps to know who confirms the stock—account manager, service department, or warehouse. A verbal "it’s in stock" is almost impossible to prove later.

Urgent delivery must also be described without gaps. Who starts shipping the same day, how it is transported, who pays for express logistics, and when the clock starts. Sometimes the part is really available, but the supplier ships it only after full payment of the invoice. For a workshop, that means one or two extra days of downtime.

Another common dispute is compatibility. The machine model alone is not enough. You need the serial number, assembly version, year of manufacture, sometimes a photo of the nameplate and setup parameters. If the service selects parts using this data, it is better to state that in the contract. If the supplier makes a mistake, they should replace the part at their own expense.

If a company supplies equipment and services it at the same time, as EAST CNC does in Kazakhstan, the spare parts section should be tied to service in advance. Then delivery and repair do not live separately on paper, and the workshop immediately knows who is responsible for selection, availability, and lead times.

How to arrange dispatch conditions

Many disputes begin not during the repair, but in the first hour after the breakdown. If the contract does not specify the exact dispatch procedure, the workshop loses time on calls, approvals, and finding someone to blame.

It is better to name two responsible people on each side right away. On the customer’s side, this is the person who submits the request and the person who grants the engineer access to the site. On the service side, you need not a general company phone number, but a specific request intake channel: phone, email, messenger, or service system. Otherwise, the request gets stuck, and no one counts the time.

What to write directly in the contract

You need to fix the moment from which the dispatch time starts. A convenient scheme is simple: the service confirms the failure based on the description, photos, video, or remote diagnostics, and from that hour the countdown begins. Do not confuse this with the first response time. Answering a request in 30 minutes and dispatching an engineer in two days are different things.

List expenses as a separate item:

• travel to and from the site;
• accommodation if the repair takes more than one day;
• daily allowances, if applicable;
• who pays for urgent dispatch outside the schedule;
• what is included in the price and what is billed separately.

Without this, the estimate can grow after the engineer arrives. For businesses in Kazakhstan and neighboring countries, this is especially sensitive: the site may be far from a major city, and travel can take longer than the repair itself.

Another disputed point is remote sites. Check how the service calculates distant locations: by kilometers, by travel time, by service zone, or by actual tickets. If the workshop is not in the regional center, this significantly changes both the price and the dispatch time.

Where time is lost most often

At secured and restricted sites, downtime often starts with access, not with the failure itself. Clarify in advance who arranges access, how long it takes, what documents the engineer needs, and whether they can bring a laptop, measuring tools, and spare parts. If security only works on weekdays until 6:00 p.m., a night visit means little on paper.

It is also useful to plan for exceptional cases. If the engineer arrived but was not allowed onto the site, who pays for the trip? If a second visit is needed because access was not granted, who covers that cost? While the machine is working, these points seem minor. On the day of the breakdown, they quickly become the main issue.

How to read the contract step by step

A long contract can easily lull you into complacency. Because of that, people more often miss not the price, but the lines that later turn into extra hours or days of downtime. That is why it is better to read a service contract by sections, not straight through.

First, put all the timelines into one table: response time to a call, remote diagnostics time, engineer dispatch time, spare parts lead time, and restart time after repair. When everything is in one place, it becomes immediately clear where the obligation is specific and where it can be stretched with wording like "if possible".

Then check five things:

1. Exclusions and extra charges. The base rate often does not include night dispatch, weekends, travel, engineer accommodation, or repeat startup.
2. The list of work and your typical failures. If sensors, the spindle assembly, or the lubrication system fail often, check whether diagnostics and replacement of these assemblies are included.
3. The spare parts section. It is important not only to know whether there is a warehouse, but also who reserves the parts, how the lead time is calculated, and who pays for urgent delivery.
4. The engineer dispatch terms. Is an advance payment needed, how many hours count as working time, who arranges access to the site, and what happens if the engineer arrives and the machine was not prepared.
5. Restart after repair. Sometimes the repair itself is included in the contract, but commissioning and the final check are billed separately.

Then ask the supplier to show the calculation for two emergency scenarios. The first is a drive failure on a workday. The second is a breakdown on Friday evening when the required part has to be ordered. These examples quickly show both the real response time and the final amount.

If you only hear general promises instead of calculations, it is better to revise the contract before signing.

Mistakes people make when signing

Most often, people choose a contract based on the lowest price. For procurement, that logic makes sense, but for a workshop it often ends up costing more. If the document does not specify an exact dispatch time, the machine can sit idle for two days, even if the contract itself was cheap.

Another mistake is thinking that a phone response already solves the problem. A consultation helps if the operator can check a sensor, a setting, or a simple error on their own. But when repair is needed, the response time should separate remote support from the engineer’s actual arrival. Otherwise, the supplier answers in an hour, arrives a day later, and formally breaks no rules.

The third weak point is parts. When signing, many people look only at the price of the work and forget to ask who pays for urgent delivery if the needed item is not in stock. Then the engineer quickly finds the faulty assembly, but the part is shipped separately, the customer pays for transport, and downtime grows every day.

Vague wording also works poorly. Words like "promptly", "as soon as possible", and "by mutual agreement" do not protect production. You need numbers: how many hours are allowed for the first response, how many hours or days for dispatch, when the service issues its conclusion, when it approves the estimate, and how long it takes to start the repair.

And one more common mistake: only the purchasing manager or director reads the contract. Without the section supervisor, mechanic, or shift manager, it is easy to miss the real working conditions. They are the ones who know when the machine usually stops, whether some problems can be handled remotely, how long engineer access takes, and who meets the service team.

Before signing, it is useful to check five questions:

• from what moment does the response time start—from a call, an email request, or confirmed diagnostics;
• when should the engineer be on site, not just available by phone;
• who pays for urgent delivery and customs charges for parts;
• which deadlines are written in hours and days, not general words;
• who in the workshop has already read the contract and agrees with it.

A good contract is recognized not by a nice price, but by how few reasons for dispute it leaves on the day of a breakdown.

A workshop example

In the morning, a CNC lathe stopped mid-cycle. The cause was not the most serious, but it was unpleasant: a sensor stopped reading the position of the assembly correctly, and the machine went into an error state. Workpieces were already in process, the shift was on site, and output stopped almost immediately.

The department had two service options. The first contract looked attractive on price. It promised a quick response by phone or messenger, for example within 30–60 minutes. On paper, that sounds good, but the engineer dispatch was in a different clause: up to 3 business days after the request was confirmed. In practice, the technician replies quickly, asks for a photo of the error, gives basic checks, but the machine still waits for a real inspection.

The second contract cost more, but it specified engineer dispatch the same day if the breakdown stopped production. That is what actually solves the problem. The engineer arrived, checked the sensor circuit, replaced the faulty component, cleared the alarm, and got the machine back into operation before the end of the day.

The price difference between the two contracts was smaller than the loss from one shift. If a workshop loses 8 hours on one heavily loaded machine, the costs are not only lost output. Add shift wages, delayed shipment, operator downtime, and the risk of missing the next order. Even with a modest estimate, one such shutdown can easily eat up the savings from the initial discount.

That is why a service contract is usually read not by the first number in the invoice, but by two lines: when the engineer will actually arrive and whether the service has access to the required spare parts. A quick response without a visit often only soothes people in words. When the machine is stopped, the important thing is not the chat—it is the time to repair.

For CNC machine maintenance in Kazakhstan, this is especially noticeable: if the dispatch terms are vague, even a minor breakdown can stretch into several days.

Short checklist before signing

Before signing the contract, do not look at the total amount—look at what will happen on the day of the breakdown. The document should answer one simple question directly: who will start solving the problem, when, and at whose expense.

If the text contains too many vague words, that is a bad sign. Phrases like "within a reasonable time" or "by mutual agreement" almost always work against the machine owner.

Check a few items:

• the response time is given in hours, not just as "promptly";
• there is a separate channel for emergency requests: phone, email, messenger, and the responsible person;
• the engineer dispatch time is fixed precisely, with a clear starting point;
• the list of work is clear: does it include diagnostics, adjustment, assembly replacement, restart after repair, and accuracy checks;
• there is a parts procedure: who keeps stock, which items are usually available, how long to wait for delivery, and whether approved alternatives can be used;
• travel costs are spelled out in advance, with no surprises for tickets, accommodation, daily allowances, and weekend visits.

There is one more small thing that often wastes time. Check who makes the repair decision if your shift supervisor is unavailable. It is better to name both a primary and backup contact on the customer side right away.

If even one point cannot be understood on the first read, ask for changes before signing. Verbal promises do not help on the day of a breakdown.

What to do next

First, pull up the shutdown log for the past year. If there is no log, gather data from shift reports, supervisor chats, and service notes. You do not need general complaints—you need repeating causes: sensor failure, drive error, hydraulic leak, spindle problems, long waits for a part.

Such a list is quickly eye-opening. After that, the contract no longer feels like a formality, because it becomes clear where the workshop loses hours and who should cover these risks.

Next, ask all suppliers for the same set of terms. Otherwise, comparison loses its meaning: one counts only dispatch, another includes remote diagnostics, and a third does not take on part of the work.

Usually, four points are enough for comparison:

• response time for remote support and dispatch;
• the exact list of work under the contract and a separate list of exclusions;
• availability of critical spare parts and their lead times;
• engineer dispatch terms: who pays for travel, accommodation, weekends, and urgent callouts.

Then look not at the annual price, but at the time to get the machine running again. That is the main indicator. A contract with a lower amount can easily lose if the engineer arrives in two days and the needed assembly takes a week to arrive. One lost working day often costs more than the difference between two contracts.

If you have several machines, make a simple table. In one column, list the typical failure; in the second, how many hours it took to restart last time; in the third, which contract term could reduce that downtime. That makes it easier to see that for CNC machine maintenance, specific terms matter—not pretty wording.

If you are looking for a supplier that covers the full cycle, it is useful to discuss machine selection, delivery, commissioning, and service together from the start. For example, EAST CNC in Kazakhstan works exactly in this format: the company supplies CNC lathes and other metalworking solutions, and also handles commissioning and maintenance. In that case, it is easier to agree in advance on the response time, the list of consumables, and the engineer dispatch procedure for your production schedule.

Before agreeing, ask for the final draft of the contract in writing. Mark the response times, the list of work, the list of critical parts, and the dispatch procedure. If this data is not on paper, the dispute will begin again at the first shutdown.