Blog article
See all stories »

Getting the basics right: Achieving 99.99% uptime and no data leaks

It seems as though tough times are ahead for eCommerce merchants. The potential for a major downturn has many battening down the hatches in anticipation of lean profits and higher expenses.

During this time, it is therefore vital that companies get the basics right. There are two dangerous temptations here: the first is to invest in flashy new systems, in an attempt to stand out from the competition, the second is to cut costs as far as they will go by using budget providers for essential systems like payments. Neither plan is exactly wrong – it’s possible that new technology can revolutionise a business and it’s equally possible that you can find a better deal for the same level of service. It is far more likely however that the best path is to go back to basics, making sure that your payment systems are efficient, robust and secure.

Uptime, downtime and payments

The average payments company has a 99.9% uptime. This seems impressive, but in practice it means that your site will be unable to take payments for 43 minutes every month. For a very small site, or one that relies on small numbers of very large payments, this might not be a problem, but for sites that rely on continually doing business that 43 minutes can translate to thousands or even tens of thousands in sales and potentially lost customers.

Planned maintenance is an inevitable part of using any digital systems – even major websites like Amazon or Facebook will have times when at least part of their huge customer base is unable to access their services, although these will be spread out geographically so that their entire userbase isn’t knocked offline at once. Any even halfway competent payments company should give their customers plenty of notice of any planned maintenance.

The majority of companies (82%) have experienced an incident of unplanned downtime in the last three years, and many have suffered two or more. There are multiple causes for this, but the most common is human error – a single typo in a command prompt or misconfigured component can cause entire networks to go offline for hours until the problem is solved. Mandatory updates from card schemes happen at least twice a year, and these can cause downtime due to bugs or clashing configurations. There are also deliberate attacks such as Distributed Denial of Service attacks that can cause an individual merchant or the payment system they use to go offline or at least slow down to the point that they are unusable. This is a particular problem when the individual components that merchants use are also composed of multiple connected cloud components – a merchant’s site might go offline because of a DDOS attack on a single service that performs a single critical function for their payment processor. Given the range of things that can go wrong, it’s surprising that 99.9% uptime is the average.

How to maintain a 99.99% uptime

Suppose we could cut downtime from 43 minutes per month to just four minutes – 99.99% uptime – while not compromising on other services. It’s something that only premium brands in the payments landscape can offer, where other companies might cut corners to keep costs low. This figure doesn’t include planned maintenance, or unplanned events such as card schemes going offline, but it is still significantly better than the majority of the market.

The first way to maintain 99.99% uptime is by creating every merchant’s payment system on a bespoke basis. That means that each merchant is running a unique product and, aside from the many other advantages, not having all clients’ eggs in one basket. Whenever possible, it is possible to make small, a la carte updates when needed them, and preferably use hotfixes that don’t cause downtime.

From there, there are dozens of ways to achieve higher uptime. Here are just a few:

  • Not being dependent on third parties, allowing standardisation of systems and fewer dependencies on variables that can’t be controlled. This allows change requests and updates to be standardised across the system – every change made is applied to the entire portfolio, avoiding complex testing and excessive updates.
  • Legacy processors are not agile with their testing due to capacity and customisation issues (customisations are problematic as they create a non-standardised environment, and therefore require exceptions for testing and increased implementation time)
  • Payment acceptance processing capabilities have redundant servers that can be moved between in real time.
  • By having APIs for everything, it means significantly reduced human error. For example, MID generation is automated via API, so it only takes five to ten minutes and avoids human errors.
  • In the instance that Visa or Mastercard do go down, alternative payment methods are provided to ensure users can still transact.
  • Constant investing in product development, maintenance and optimisation to support the never-ending demands of the market.

The importance of choosing the right payments partner

Considering how much can be lost to downtime, it is vital that merchants start looking at how they can increase uptime and thereby keep their customers from seeing the dreaded ‘check back later’ screen. The chances are that unless you have a deep, longstanding relationship with a customer or sell such a unique product or service that they can’t go elsewhere, then as soon as they see that your payment system is down, they won’t wait. Why should they when another site is just a google search away?

Think of your eCommerce site as a luxury brand, even if your branding is anything but. At first glance, luxury brands aren’t much different from their mass-market equivalents, but you’re paying for craftsmanship that will mean that £100 boots last far longer than £50 boots. Payments systems are the same, but with the key difference that well-crafted systems aren’t significantly more expensive than those that cut corners. You can find a payments partner willing to create bespoke products that offer high uptime and other sought-after qualities without breaking the bank, and that will be critically important at a time when merchants will struggle.

6642

Comments: (0)

Scott Dawson

Scott Dawson

Head of Sales & Strategic Partnerships

DECTA

Member since

21 Apr 2023

Location

London

Blog posts

8

This post is from a series of posts in the group:

Outsourcing Banking, Financial Services Operations

Group is focused on accelerating Business Process Transformation through the use of the latest Fintech point solutions, AI, RPA and Design Thinking.


See all

Now hiring