The business proposal that never got off the ground

I wanted to share a proposal I worked on in late 2013. It’s not particularly compelling, nor did it get very far. But I wanted to share it as an example for anyone who is working on their first business plan or idea.

The idea wasn’t particularly new, inventive or unique. It seems pretty obvious in-fact; I wanted to reduce uncertainty for people when they were booking airline tickets.

Complexity is costly

Way back when I was at University, one of the subjects I studied was behavioral economics - a fascinating combination of psychology and economics; I now suggest to people that behavioral economics is the missing link between reality and theoretical economic models.

Behavioral economics studies the effects of psychological, social, cognitive, and emotional factors on the economic decisions of individuals and their consequences.

I have written previously on complexity in product offerings and consumer choice in Private Health insurance. Probably the better known example, however, is in telecommunications.

In Australia, the telecommunications industry adheres to a voluntary code of conduct. We’ve had various codes of conduct for the telecommunications industry in Australia since 2000 (although it was written into legislation that the industry develop a code of conduct, so perhaps it’s not surprising).

The current code, among other things, ensures that the plans that are sold to consumers are:

communicated in a way which is clear, accurate and not misleading, to allow Consumers to make informed choices (Section 4.1.1).

I strongly agree that in order to make an informed choice consumers need to be shown the options in a clear, accurate and honest manner.

Now if you’re wondering about how complexity might affect consumers in air-fare pricing, let me quote an example from ITA software back in 2001. Let’s use the example of flying from San Francisco to Boston - two airports that are relatively close.

If you restrict it to same day flights that have at least two stop-overs, you would have had approximately 2,000 options to select between.

If you are willing to consider all flights that get you to your destination the same or next day, there are over 1 billion possible flight paths. Just for San Francisco to Boston.

Now obviously when you go to book a flight, you don’t consider even a fraction of the total possibilities. So how often are you sure that you have gotten the best deal?

The business proposal: “Fair Fair”

In the proposal document, I started by posing some questions:

####Introduction

The last time you booked a flight how did you ascertain if you were getting good value?

Chances are you used an airfare search site (meta-search) such as Expedia, Kayak, Orbitz, Skyscanner, Google’s ITA Travel Matrix, Hipmunk or another site. All of these sites share a common operating model - show the traveller the most relevant fare around the date they want to travel and collect a commission on the booking (if the traveller books).

How many sites did you check and how many total searches did you perform before you ended up booking your flight? How confident were you that you got the best fare? Do you know when the best time to book is?

I also wanted to highlight, as succinctly as possible, why the current method could be improved. I had talked to a few people in the industry, and a representative from a large “Global Distribution Company” had commented on the ratio of flight searches to flight bookings. I was surprised by how high the ratio was.

Would it surprise you to learn that the proportion of searches performed to completed bookings can be as high as 350:1 for leisure travellers and 100:1 for business travellers?

There is a clear pain point here for travellers around the uncertainty of flight prices.

Fair Fare would seek to reduce traveller uncertainty, leading to increased conversion to booking.

Next I outlined what I was seeking; I had invested significant time in the idea - and had already developed working relationships with two Global Distribution System (GDS) companies.

GDS companies are the market makers that sit between the airline and the retailer/travel agency/online booking service. They aggregate flight availability, pricing and other options and make it available on a commercial basis.

In order to understand the market, one would need access to a significant amount of pricing data. I had been able to get trial access to two of the largest GDS platforms - but these were setup more for travel agents or booking websites - not so much a big-data analytical exercise.

In order to access any significant quantity of pricing data, however, I would need to either fund it myself or seek an investor.

Proposal

Felix Barbalet is seeking Angel funding to conduct the initial phase of development that involves acquiring 1 years worth of data and developing the IP for flight price analytics at scale.

This would be a 6-12 month proof of concept to develop IP and proove the concept. Following phase 1, Fair Fare could either seek funding to take the idea to market or seek a buyer for the IP assets.

Felix has been working on the idea for about 2 months and is getting to the stage where validation and funding are required to move forward, primarily due to the cost of acquiring flight pricing data at scale.

In those two months Felix has developed relationships with two of the worlds largest Global Distribution Systems (GDS) - REDACTED and REDACTED. Felix now has access to the developer portals of both companies, but for access to data at any sort of volume, funding is required.

Next I went into a more formal business model - which I had developed using the excellent Business Model Canvas.

Draft business model

1. Customer segments - who are we creating value for?

Price sensitive leisure travellers

• Travellers aged 30 and under (Millennials)
• Expecting better ‘comparison shopping experiences’, used to mobile transacting
• Travellers aged over 30 - more price sensitive

Price sensitive business travellers

Travel agents

• Travel agents will typically have access to fare price predictions but only at a granular level
• Fair Fare would provide real-time access to detailed, low level predictions for each booking

Online travel agents

• Using the Fair Fare API, OTAs could provide predictions on their sites

2. Value Propositions - what value do we deliver to the customer?

Fair Fare would deliver value to the customer by reducing or removing uncertainty around their purchase of flights.

3. Channels - how do we reach our Customer Segments?

Direct- online desktop/mobile website Direct- mobile native apps Indirect- online (Resale/APIs)

4. Customer Relationships - what type of relationships do our Customer Segments expect?

Fair Fare would seek to develop a ‘Trusted Advisor’ identity - saving travellers time and money - and in doing so become the destination of choice for savvy travellers and increasing the conversion rates to booking.

5. Revenue streams - what are our customers willing to pay for?

Advertising

Eyeballs on the site - as the Trusted Advisor for flight price predictions, Fair Fare should generate significant traffic volumes.

Subscriptions (Freemium model)

For business / travel agents or others who might want more detailed pricing data.

Referrals to booking sites

For example Expedia pays $3 per referral that results in a booking.

Selling insurance against adverse price movements

With appropriate actuarial models, Fair Fare could offer insurance to travellers against adverse flight movements. For example, “book now through Fair Fare Insureaprice and we Guarantee you’ve locked in the best price for 30 days”.

Acquisition for IP/technology assets

If Fair Fare accuracy was better than competitors, (or even on-par) it could be an attractive acquisition target for the IP/Technology assets.

6. Key resources - what key resources are required to deliver on our value proposition?

Physical

• Compute infrastructure

Intellectual

• Data: fare price and capacity utilisation data at a large scale.

• IP: Predictive and analytical models

Human

• Appropriately skilled data scientist to work on the data in phase one (initial IP development).

Financial

• Both data and compute infrastructure

7. Key activities - what key activities do our value proposition require?

• Data processing at scale (problem solving)
• Predictive accuracy (IP development)
• Delivery channels to get to market (Platform / network)

8. Who are our key partners?

In Phase 1 a key partner will be the GDS/data provider. Another key partner would be the infrastructure provider. In later phases there would be more key partners.

9. Cost structure inherent in business model

There is a significant upfront fixed cost for acquiring the data, plus variable costs for processing and responding to user queries; the business model relies on economies of scale to realise a profit.

We are working on getting exact costings from the two Global Distribution Systems we have built relationships with, but at this stage it looks like 1 years worth of data at a large scale would be 40-50 thousand dollars. Infrastructure costs are likely to be on the same order of magnitude.

Other costs in phase one are minimal (Felix is not seeking funding for a salary or other expenses during the IP development/proof of concept phase).

Depending on the strategy following phase one, there would be further costs to take the product to market.

10. Competition

There are companies operating in this space already - namely Kayak and Bing.

Kayak and Bing offer a fare prediction service currently (in the US and UK markets only), following acquisitions of startups Farecast and SideStep respectively.

I am not aware of any start ups which are pursuing this concept, but of course there are many startups in the travel space. Notable ones include Hipmunk, Rome2Rio (Melbourne based) and FlightFox.

Lessons learned

In hindsight there are plenty of reasons why this proposal didn’t go anywhere. I wasn’t particularly aggressive at seeking funding, and shortly thereafter I took on a number of consulting contracts, which resulted in much less time for development of this idea, and the proposal itself wasn’t very appealing to an investor - high risk with uncertain return - and many parts of the business model lacking in detail.

It was interesting, none the less, and I learned plenty about the industry and air fare pricing.

What prompted me to write about it today however, was this article on The Huffington Post - which details Google Flights - which looks to be the evolution of ITA Travel Matrix. Google Flights looks pretty good (not surprising, this is a Google sized problem). I’m not sure what Google gets out of it - they don’t let you book fares through any of their sites so I don’t think they are collecting commissions.

If I was working on the idea today, I wouldn’t feel at all discouraged by the thought of competing with Google (or any of the other sites). I would do more work on getting a prototype together and instead of a written proposal, I’d put together a slide deck. I wouldn’t give up so quickly. And I would talk to everyone I could about it.

If you’re just getting started on an idea, keep going. Talk to your friends and family. Talk to your colleagues. Talk to those in the industry. Blog about it. Put it aside for a break. But don’t put it aside for too long - keep working on it!