skip to Main Content

The Four Pillars for Optimising Online Conversion (Part 1)

four pillars availability performanceIn the world of travel, the factors that can influence getting a booking or not in the B2B world are related but ultimately different to those in the B2C website world.

At Triometric we have identified 4 key areas of influence in the conversion process for distributors and suppliers. Each pillar plays a crucial role in the business of selling travel products through intermediaries, be it a hotel room, an airline seat or a related ancillary such as a hire car or an excursion. In much the same way as a chair has four stabilising legs to stay upright, the same argument can be made for an integrated and holistic view of the key elements involved in selling travel products online, especially through distribution partners via a network of APIs.  In Part 1 we focused on Availability and Performance – the basics for a distribution system, in Part 2 we discuss the importance of context in terms of ensuring the Relevance of offers made and at the best Price for that conversion at that time.

How important are these pillars?

globe book four pillars blogYou could argue they apply to all e-commerce selling, but they resonate in particular with travel companies operating in a highly competitive and fragmented online B2B environment. At Triometric we believe these four key areas are the cornerstones of online travel distribution (and by implication conversion). The ability to monitor and measure them is the difference between surviving and thriving in a very dynamic environment.

So what are these essential pillars?

The overall objective behind these pillars to deliver the right product to the right person at the right price.  In a nutshell, the pillars are:

  1. Performance – of servers, applications, networks, APIs
  2. Availability – insights into how inventory demand is meeting available supplies
  3. Pricing – insights into the dynamic changes influencing offers and bookings
  4. Relevance –insights into customer buying context to present the best products

Collectively they represent a journey from considering IT issues through to getting the right offer in front of a prospective customer by understanding more about.  And that does not mean analytics to more forcefully sell the same old product. It’s about understanding the customer and their needs.

This article focuses on the first two pillars – the two that together are the basics for a distribution system.

Pillar 1 — Performance

In the world of online travel, technology is the plumbing through which the critical business data flows between distribution partners.  The API platform is the engine of the business and it needs constant monitoring. Tuning into cues like uptime, speed, connection issues and usage only provides the basic IT performance indicators need to detect and respond to the issues that impact the system on an hourly basis.

What makes the difference is attention to and having ‘the detail’ – average statistics really don’t help.  For example if average response times go up (not good) what do you do?  If the answer isn’t overwhelmingly obvious e.g. some broken server in the server farm then you need lots of decent operational data mixed with business level information. The process is not just, whose traffic is causing the demand? , but also, what products are they searching for?  Oh, and which ones are a problem? So for airlines, you would need to know what the average response time is for all of the key routes or per if you are in the hospitality sector, all the top destinations or specific properties. Poor responses times means less page 1 search result outcomes with the prospective traveller and that means lost revenues.

Another example from our impatient world of travellers is timeouts.  How many (milli)seconds do we have to respond in before our online agent determines they can’t wait any longer and simply goes with another faster supplier or distributor’s offerings? How much of that is due to network delays or my servers? What proportion of timed out requests impact by top products?  What is that costing my business?

Remember that timeouts are an IT concept whilst products are a business concept.  The IT data is gathered by monitoring the IT system whilst getting to the business breakdown requires you to process the data content in the request and/or response. To be effective with addressing the issues within our first pillar you need to tie the two together.

Pillar 2 – Availability

In our availability pillar we are considering the problem of what product offers our API is consistently   making against the ever changing travel product search patterns arriving at our front door.  If we are a distributor our concerns will be around our direct sales contracts or third party distribution contracts that determine our level of inventory for a particular location or route across all of our suppliers.  If we are one of the suppliers, then our concern is around capacity – how many rooms in a location or seats on a route – which we have or have enabled for sale through a particular distributor, often at a particular rate.

To manage availability, we need to be constantly measuring what we actually send back in our responses.  Mistakes with the mapping of product against search criteria e.g. London, GB vs. London, UK or not enabling fare buckets will immediately impact booking revenues.

There are couple of ways of reporting on availability.  I can either track what proportion of requests result in a response offering at least one product (% Availability) or I can track how many product offers whether that be room options or flight departures are being made per request (Supply-to-Demand).

Availability reporting isn’t just about products either – an agent consistently searching an API for products with low or zero availability is loading the API system without driving revenue.  This presents itself as either an opportunity to offer new product or to work with the agent to limit platform load.

There is a natural presumption that we are addressing each pillar in turn so reports of a low %Availability value should be able to factor out IT related (pillar 1) issues.  If it is still low then a very detailed breakdown of the product being searched is going to help – this includes check-in and other possible restrictions on particular product.

Supply-to-Demand reporting provides the opportunity to track the level of choice that we are offering the traveller.  Reductions in Supply-to-Demand indicate a drop in consumer choice and a predictive shortening of available inventory perhaps across certain travel dates.  For distributors, this is the time to engage their contracting skills with new suppliers or negotiate more inventory with existing suppliers.  For the supplier, this suggests a high demand for product they have either failed to offer/enable or there is a genuine lack of capacity.   Adding capacity obviously means larger aircraft, more flights per day or a new hotel depending upon the business in question.

The Time Dimension

All of our pillars are about having the right information but there is also a significant issue around timeliness of data.  So often organisations rely on overnight processing of log data which is intrusive to collect and its processing is delayed to avoid further booking platform loading. What is needed is real-time element to the reporting.

Finding out about a revenue impacting problem that happened yesterday is irritating.  Having that problem 365 days a year is expensive.

Our next blog in this series will address the remaining two pillars around Revenue and Relevance

Back To Top