Service Level Agreements – SLAs Explained: Controlling and Measuring Good Service

Provide services, ensure quality and performance, and at the same time neither pay too much nor forego relevant content. With the right SLAs, this becomes possible!

SLA as a term, everyone has probably heard it before. Often one has the impression that this is exclusively about documents with the same cryptic depth as general terms and conditions or data protection declarations. But SLAs are the part of every service that should really be understood.

Service Level Agreements – that sounds a lot like just another dry document. But behind the classic SLA are many more elements and information that can be interesting for any type and form of service. Perhaps you have already encountered SLAs when you signed a contract for web hosting or a web server. There, the topic is usually dealt with under the keyword “availability”, and you get the impression that it is mainly about securing this accordingly. But once you understand the possibilities of SLAs, you quickly learn to appreciate the details of this modern form of service agreements!

Meaning of SLA – Service Level Agreements explained

Let’s just start at the beginning: SLAs are always relevant when it comes to providing a service. In most cases, these are not simple services, but permanent, so-called ongoing services. These are not just called up a few times a year, but are provided non-stop or over a long period of time – Internet services are a good example. In our world, where being offline is interpreted as being “cut off from the world” and messenger or email have become the number one communication tool, there is a corresponding expectation of constant availability. But nothing really works without interruption, no service can be provided non-stop: Even in a restaurant that is open 24×7, the kitchen has to be cleaned, an update has to be applied to an email server in the company, or a service simply fails from time to time. People get sick, tools break down and machines inevitably have errors.

Precisely because services and the provision of services cannot be without interruption, the aim of the SLA is to create maximum transparency for both sides – i.e. the provider and recipient of the service. If there is an interruption of one hour per month at your email provider, this may be completely normal and legitimate within the scope of the agreements. Especially, of course, if this interruption is planned and announced, and ideally takes place at night or in a time window where you, the customer, don’t notice it at all. It is precisely this subtlety of assurance of availability that is at stake – and knowing it brings decisive advantages for a multitude of services that we consume every day, consider for commissioning and, if necessary, also want to cancel. After all, what’s the point of paying for or ordering more than we actually need?

Ensuring the essence of the service via SLA – delivered as ordered

Satisfied customers are generally found wherever they get exactly what they ordered or expected. Basically, this requires a few elements that form the basis of every service – regardless of whether the supplier is an online service, a pizza service or a car repair shop:

1. Dlear definitions and designations

The recognizable, standardized designation of the delivery item.

A pizza is not a sandwich. The customer wants to be delivered what they already know – they also say what is “common”. Most services in today’s world are standardized, even if it doesn’t look that way. But everyone has a certain idea when they think of a dish and tap the delivery app button. Therefore, in case of dispute, it is actually assumed that certain things are not common or even forbidden: For example, a pizza served on waste paper would violate food hygiene and a non-carbonated cola would not correspond to the relevant known soft drink. But the point is not to make everything the same and indistinguishable, but to avoid discussions on familiar terms. After all, if everyone freely defines everything as a pizza, no one will know what a pizza is anymore.

2. Relevant distinctions and details

A tire change is not a wheel change, an inspection is not a repair job.

Small differences, big impact: If you drop off your car with the summer wheels, but get the same wheels back with freshly mounted winter tires, you would probably be very surprised – that’s why some terms are important when specifying a service. After all, a wheel change refers to the combination of rim and tire, while a tire change refers only to the rubber. However, interpersonal trust and cooperation remain important, because if the workshop you trust asks you after an inspection whether the broken cover on the seat adjuster should be replaced for 5 euros, a pragmatic approach and “cooperation” would certainly be more effective than a repair recommendation with a follow-up appointment – because the new journey alone will cost more than the part itself. Communication is therefore indispensable and should lead to clear expectations on both sides right from the start, or the framework for leeway should be defined.

3. Critical delivery objects and compatibility

Just like allergens in food, missing interfaces in an online service can make for critical details.

People are allergic to ingredients like nuts – and cars to the wrong fluids. The fact that an engine should be filled with the right oil is a result of the manufacturer’s specifications. When ordering food, it can actually be life-threatening if the brittle cup in the ice cream parlor is garnished with peanuts. Here, too, special requirements should be pointed out from the outset. In the digital world, for example, not just any online store is booked, but care is taken to ensure that, among other things, payment services, warehousing and user management are compatible with the customer’s requirements.

4. Measurement, penalties and extensions

Relevant, common aspects, time scope and completion, and the possibility of reducing the price or adding to the service if the conditions change.

Whether a service is rated as good or satisfactory depends on a number of factors. The rattling car is running smoothly again, the pizza ordered was tasty – but how do you measure technical details like availability? This is where standards like “Three Nines“, which corresponds to 99.9% availability, come into play. What looks like 100% availability at first, quickly becomes the reality of about one hour of downtime in a monthly availability calculation. If this is an online service for reserving tables in a restaurant, an agreement to maintain the software every Monday night at 4h00 in the morning might be perfectly fine – but if we are talking about alarm systems and medical technology, this so-called availability level would no longer be sufficient. In such a case, a higher service level can be negotiated, which can be correspondingly expensive. Above a certain level, we then also speak of high availability and redundancy.

5. Success factors and conclusion

Acceptance criteria and result-relevant aspects that indicate the end of the service.

Here I like to take the example of the package thrown in front of the door: Everything can be perfect – the sensitive laptop was neatly packed, padded and also labeled “Fragile” and handed over to the transport service provider. But on the last mile the parcel carrier throws the box on the floor and the hard edge on the step to the entrance of the house catches a weak spot. Now the display is defective and all the effort is in vain. In the other case, if exactly this does not happen, the completion of the service would happen as follows: After 2 days of package delivery, the delivery is made as agreed, the package is handed over properly and the receipt is acknowledged. Therefore, anything that falls within this scope should be recorded accordingly for both service and item. When you buy your new refrigerator, you will certainly be happy if it is not only delivered to “curbside” but to the place of use – ideally including taking away old appliances and disposing of packaging.

Other important elements of an SLA – what you might not think of at first

Common SLAs contain numerous details, such as the responsibility of the parties. Only those who provide all the information according to what the service provider requires can assume that the result will be as desired. It is therefore also important to know which standards must be adhered to and how escalation paths are designed. What can go wrong, will go wrong – and it should be clear with whom deviations should be discussed. In the area of quality management, this would be a factor of customer satisfaction at the same time, because even in the worst case scenario, an unsatisfied customer can usually be satisfied again with a service saved by escalation. In the end, it is something else whether one is compensated with a reparation or a discount when a delivery is delayed or not even an apology is issued. Finally, this also shows that the service provider is aware of his duty and takes the customer seriously.

A fundamental part is the issue of pricing. I like to use the comparison of an ice-cold can of Coke in the Sahara: If money is no object, you can open an ice-cold can at any time and refresh yourself. Sure, it would probably be the most expensive can of Coke to consume. For this reason, the price of a delivery will be linked to basic conditions, as we can read in the instructions of any parcel service provider: No islands or war zones, mostly exclusively mapped areas with postal addresses. For everything else, there are courier services that, for example, fly a component for a Formula 1 racing car overnight by private jet from the German manufacturer to the race track in Spain. And that can be the price of a cold fizzy drink in the desert, but it can also save a racing team with a budget in the billions.

In the end, every expectation, every normal case, and every deviation to be avoided should be well documented – whether it’s contacts, reports, and quantities, or finances and compensation in case of non-performance.

OLAs, UCs and service tiering – the next level of an SLA

My passion for service delivery has taken me all the way into the internal services of companies. Following the principle of “build every service so that it could survive in the market”, I have always aligned IT support, customer service and information content in such a way that, even when used purely internally, it was clearly aimed at satisfying (in this case, internal) customers. The reason is quite banal: Good service is fun! Anyone who has a friendly service employee on the phone immediately gets in a better mood. And any service that is designed with passion and a clear focus on current needs generates satisfied customers. SLAs for services within companies are mapped in so-called OLAs. A well-positioned IT department knows exactly when its first internal customers boot up their computers in the morning. Then they also know that Mr. Schmidt always forgets his password after his long vacation, so IT support is available early in the morning. Or the online store on Black Friday, which likes to go down to its knees when thousands of new customers order for the first time and the nerves of the marketing department are on edge. This is where dynamic agreements make sense – because getting more power for high-load phases can be crucial to success.

So remember OLAs as clear service offerings within companies that are easy to understand. That’s where UCs, or underpinning contracts, come in – they’re effectively subordinate contracts. Does production care about the SLAs that IT has with supplier XY? No, and that’s why it’s the IT department’s job to set up its own supplier relationships and SLAs for e.g. spare parts deliveries in such a way that the internal OLAs work. Example: If the failure of a printer in the production line has to be repaired within an hour, it would be good if the internal IT not only had a replacement printer at hand (keyword: availability), but above all an SLA with the maintenance service of the printer manufacturer. And this SLA should have a response time of just a few hours. That seems too expensive? That’s exactly when service tiering comes into play – namely how critical a service is, whether you provide it yourself, have it done externally, or even mission-critically go both ways. Do you have a tire service and only one machine to balance wheels? Then you should consider how much loss a breakdown causes and what, on the other hand, the purchase of a second machine or the arrangement of a readily available maintenance service costs. But be careful: not every service has to be classified as critical and it is even possible to include seasonal factors – in the tire changing season the issue is much more important than in the first weeks of January when mainly new batteries are sold because many cars do not start.

Where is the best place to start?

So there are a lot of details and new terms when you first come into contact with the topic of SLA. However, as we have seen, the topic not only deals with aspects that are contractual and legal in nature, but also focuses on technical details. The recommendation is therefore ideally to talk directly to the departments and experts in your own company. In most cases, the topic of services is completely different if you look at it from more than just a financial perspective. If you focus on the details of the services and what is ultimately decisive for success, you are at the beginning of a truly resilient service relationship that is satisfactory for both sides.

Philipp Schneidenbach ist Experte für Governance, Risk und Compliance, Enterprise Architecture und IT Service Management. In seiner Rolle als Principal bei einer Münchner Unternehmensberatung trägt er auch die Verantwortung für Legal und Compliance sowie Digitale Transformation. Philipp teilt seine Insights als Sprecher auf internationalen Events und hier bei MoreThanDigital.

Comments are closed.