Picture a retail chain or online seller facing a familiar problem: for three or four months of the year — holidays, back-to-school, seasonal sales — the load on their systems triples or quadruples. To handle those peaks, the company builds a fixed server infrastructure sized for maximum demand. For the remaining eight months, most of that capacity sits idle, yet the electricity bills, maintenance contracts, and licensing fees keep coming. Many Turkish SMEs walk into this trap without doing the full calculation, and a few years later they find themselves stuck with a system that is both undersized at the wrong moments and chronically expensive at all others.
The idea behind flexible infrastructure is straightforward: if your capacity needs vary significantly across the year, why pay for a fixed capacity that matches only your busiest weeks? For most of the history of corporate computing, the honest answer was ‘there is no other way.’ Buying, racking, configuring, and maintaining servers is a capital-intensive, time-consuming process, and once the hardware is in place, adjusting it is neither quick nor cheap. Over the past few years, however, a different model has started attracting serious attention — one built around service providers who offer computing resources on a shared, pay-as-you-use basis, made practical by the steady improvement of broadband internet connections.
The underlying concept is the ASP, or Application Service Provider, model. Instead of purchasing software and hardware outright, a company uses resources hosted in a service provider’s data centre, requesting more capacity when demand rises and releasing it when demand falls. In Turkey, firms offering this kind of arrangement are still relatively few, and their services tend to be organised around specific software packages — accounting systems or ERP platforms delivered over the internet rather than installed locally. Treating the infrastructure layer itself as something elastic, scaling server capacity up and down in response to real-time demand, remains an uncommon practice rather than a standard one.
Even so, the numbers are worth examining. Consider a company with four busy months and eight quieter ones. In the first scenario, the company builds infrastructure to handle peak load; during the slow months it runs at perhaps twenty to thirty percent of capacity while paying full costs. In the second scenario, the company maintains a smaller permanent base and rents additional capacity only during the high-demand period. Fixed costs fall, variable costs appear only when genuinely needed. On paper the second scenario is clearly more attractive. The practical difficulty is that Turkey does not yet have a mature market of service providers capable of delivering this second scenario reliably and at competitive prices.
The value of flexible infrastructure goes beyond the cost comparison, though. Growth scenarios matter just as much. A company opening a new branch, launching a new product line, or receiving an unexpectedly large order needs to scale its systems quickly. With fixed infrastructure, that means procurement lead times, installation work, and budget approvals — all of which take weeks. With a flexible model, the same expansion can happen in a fraction of the time. Turkey’s strong economic growth during this period is pushing many SMEs into exactly these rapid-expansion decisions, and infrastructure agility has a direct bearing on how much those decisions end up costing.
The practical difficulties should not be understated. Data centres and managed hosting providers in Turkey are growing but have not yet reached full maturity. Relying on an external provider means accepting dependence on that provider’s technical quality, reliability, and contractual terms. Service-level agreements, data security arrangements, access guarantees, and liability in the event of downtime all require careful scrutiny before signing. Internet connectivity is another real constraint: ADSL is spreading, but corporate-grade connections with guaranteed bandwidth still carry a meaningful cost, and options outside the major cities remain limited. A flexible infrastructure model that depends on a flaky internet connection delivers neither flexibility nor reliability.
For SME managers, the decision ultimately comes down to a few concrete questions. How uneven is your workload across the year? Is your growth rate making infrastructure planning difficult? Does your internal IT team have the capacity to manage servers well, or would that effort be better spent elsewhere? The answers to these questions reveal whether a flexible, service-based model genuinely offers better value than a fixed investment. In Turkey today, the options for implementing this model fully are still limited, but the direction of the market is clear. Running the numbers now, before the market matures, means being ready to make a well-informed choice when the right moment arrives.
This article was originally written in Turkish by Gökhan MERCANOĞLU on March 26, 2007 and has been automatically translated into English and other languages using machine translation.