The Real Cost of Timeouts
In the cut throat competitive world of travel distribution– speed (and not just price) is of the essence, because several players will be lining up to compete with inventory. It really comes down to having the right inventory at the right price in the right time. All achieved with XML messages across the supply chain. If you are an inventory supplier and your offer gets timed out for whatever reason the opportunity to present an offer in that round of requests is lost for good. But the missed opportunity doesn’t stop there.
Timeouts (especially several timeouts) are signs of poor performance prompting your inventory servers to be excluded from subsequent requests for a set period. No offer on the table equals zero chance of conversion to booking. (If you want to know more about Timeouts and how they impact the travel industry check out our blog on the topic).
What is the cost to your business?
A single transaction timeout once in a while may well have negligible impact. But in reality speed of response across web services is critical and periods of deactivation are routinely used by the requestor to speed up the compilation of aggregate responses. In this environment timeouts occur often and have a much greater aggregate cost than many people in the industry realise. Let’s take a look at an example with some typical numbers to illustrate the point.
Timeout Impact in a single day
A typical average wholesaler receives 100m requests per day. His Look-to-Book ratio is 10,000 to 1 which gives him 10,000 conversions into bookings. In an average day 5% of his offers get timed out which translates into 500 lost bookings from those timeouts.
Now let’s make some financial assumptions:
— The average cost per successful transaction is £500.
— With a margin of 20% each booking contributes £100 revenue.
— From a total daily turnover of £5m of which £1m is revenue
— Losing 500 bookings due to timeouts
— Can cost the operation as much as £50,000 revenue – in one day!
But timeouts can also lead to a wholesaler being temporarily deactivated from receiving further requests from certain channels, typically for an hour, exacerbating the missed opportunities for offers and conversions. If indeed a forced timeout occurs due to slow server or network performance then other suppliers with similar timeout protocols and thresholds may also be quick to invoke timeouts.
Let’s get a sense of what the financial impact might be as result of a single hour’s worth of loss of business with this channel imposed by the earlier timeout(s).
Timeout impact when your biggest channel times you out
We’ve made the assumption that a channel represents 10% of total booking value for a typical day. With a larger customer (compared to an average customer with only 2 or 3% of daily bookings), timeouts will have noticeable repercussions.
So taking a typical 18 hour day (minus low overnight activity) the picture looks like this.
— There are around 556 bookings per hour
— As a result of a deactivation period of one hour a (because of earlier timeouts)
— 56 bookings from the preferred channel are lost
— These 56 lost bookings have a total revenue value of £5,556.
So when a timeout (or repeated timeout) occurs resulting in a boycott from one single customer the additional cost to the business is £5,556 for each and every hour the deactivation period is in place. This is clearly on top of the sum of £50,000 already lost in the 5% daily timeout losses described earlier.
If an accommodation wholesaler has 100 suppliers most of whom will have timeout protocols in place for bad connections or slow response times, then it is easy to see that the additional financial losses due to timeouts can quickly mount up. But these hidden costs of timeouts are large undetected and often overlooked.
These real-life examples from a typical accommodation wholesale operation aptly illustrate that timeouts can have serious financial ramifications well beyond the initial timeout. Being able to recognise when and how timeouts occur helps to mitigate both the timeout and the associated financial loss.
How Triometric Can Help
Of course, a server connection timeout error does little to tell you what went wrong or why the error happened: it just identifies that the error occurred. Timeout errors can happen for a number of reasons. The server, the requesting device, the network hardware, the network connection or even the application can be at fault.
This is where Triometric’s analytics platform can help. It is a proven platform for monitoring the performance of your web services and applications against key indicators. We give distribution or IT managers visibility into the performance of servers and networks. In addition, a closer analysis of XML data can identify further lost business opportunities, for example the instances when timeouts are not triggered even though responses are presented too slowly to be included.
Trio analytics platform is the perfect platform for monitoring the health of your web services and spotting timeouts as soon as they happen. It can be positioned as a key performance indicator on the dashboard which updates in near real-time and an alert email system can also be set up which can alert you when custom-set timeout SLA is transgressed even when you are not in the office. Not only can the system tell you when timeouts occur but also help identify how and where they occur, so that corrective action can be taken as soon as possible.
Response times is key to managing timeouts, which are essentially just one of the types of errors that can be recorded. How Triometric Trio analytics platform can provide monitoring and resolution in this critical aspect of online travel will be the topic for our next ‘Use Case’ example.