How to figure out where: decoding the ‘colocation versus on-premise’ library

For anyone researching ‘on-premise vs. colocation,’ the results soon become repetitive. Google, AI, and YouTube all return similar material on the pros and cons of each option. Some even include public cloud as a consideration. Yet they all reach exactly the same conclusion: that as a business, you need to figure out what is right for you.

Whilst the advice is sound, it is also limited. How exactly can a business assess which infrastructure option is right for them? We believe it comes down to four key questions.

Question 1: What have you already got?

If you are considering digital infrastructure options, it is likely that you have existing investment. This will include hardware, staff, dedicated space and support, such as power and security.

It is also critical to qualitatively assess the data itself. Not all data is created equal, and business need to ask how ‘special’ that information may be.

Consequently, businesses need not only a careful inventory, but also an assessment of what that technology delivers. The objective is to know WHY the business is considering either an expansion of existing on-premise capabilities, or a move to colocation (or even into the cloud).

This is not just limited to those companies looking to grow. Event at this initial stage a business can realise that the legacy set up is sufficiently complex that it must be kept on-premise to avoid an expensive and disruptive architecture redesign.

Question 2: What control do you need?

By control, we refer here to two separate, but related areas.

Firstly, what control over the data must a business exert? Compliance demands such as HIPAA and GDPR may mandate hosting sensitive or personal data in an owned location. Alternatively, is the data itself so commercially critical that it necessitates on-premise housing?

Secondly, there is control of access to the hardware and environment. This is not just a security issue. Fast, 24/7 contact with the hardware impacts on operational reliability and the speed of remediation.

Thirdly, there is operational control of the data and workloads. The ability to allocate the right workloads in the right place is a critical component of an active data strategy that yields the greatest return.

Question 3: How important is reliability?

Following on from the issue of access, it is important to accurately gauge what tolerance the business has to downtime (if any).

At this point, consider the various elements that can cause downtime: not just technology issues but also power disruption, overheating, adverse weather and even staff shortages.

It is worth trying to map this ‘tolerance’ onto the established Uptime Institute classifications to get an idea of what reliability your business demands.

Question 4: What are your growth plans?

This is the most important question on this list. Defining the plans for the growth of the business and understanding the impact on that businesses technology, is critical.

It is all but guaranteed that any business will see a substantial increase in data. However, there may also be other sources of growth: increased staff, more operational sites, or moves into new markets or geographies.

These latter concerns also force a focus on the connectivity within the facility. How well and how quickly can a data facility in one country, region, or locality, serve a market in another?

At this point it is also worth considering if growth plans will also include the use of artificial intelligence (AI) and the Internet of Things (IoT). The impact of AI on digital infrastructure is an area of growing concern for CDOs because of the huge demands AI places on hardware. Meanwhile, IoT devices are currently forecasted to reach seventy-five billion globally by 2025[1].

Conclusion: Pulling it all together

The above questions help to define the return seen on investment into various digital infrastructure options. Historically, decisions have been made on the financial basis of the lower capex offered by colocation. But as the above operational considerations show, the ‘cost’ in total cost of ownership, needs to go far beyond initial spending.

Decisions on the correct digital infrastructure for a business should be grounded in an accurate assessment of the current technology held, compared to future business objectives. Remember that as these objectives change, and the business shifts, the right answer today may need reviewing tomorrow.

To see what option suits you better, try out our new Digital Infrastructure Assessment Tool, designed to point you towards the best digital option for your developing business.

-> Try our new assessment tool

[1] See Internet of Things (IoT) connected devices from 2015 to 2025 (in billions)  | Download Scientific Diagram (researchgate.net)