"Big Data" has evolved from a marketing buzz-phrase to a marketing cliché over recent years. But brands still have a way to go before they understand, let alone fully utilise, the potential of the datasets available to them. That was the overriding message of Blind Data, an event organised by UK commercial TV trade body Thinkbox and held in London this morning.

A view from the client-side came from Peter Duffy, marketing director at easyJet, the low-cost airline, who offered a pretty compelling case study showing how the company is using a mix of its own data and sets from external sources to optimise its media planning. And there's no reason why many of the lessons from easyJet's story aren't applicable to brands in other categories.

easyJet is a bigger company than most (it flies 60m people a year) but, up until recently, it did not have a big ad presence in broadcast media, spending just £200,000 on TV channel across the whole of 2010. But it now spends well into the seven figures on TV each year. This is part of a broader process: three years ago, 70% of the overall media budget went on direct response and 30% on branding; by 2013, this ratio had flipped.

The results of this shift towards branding are impressive: year on year, there has been a 5% increase in the value of easyJet tickets sold by paid media, even as the company's overall media spend has been reduced by 15%. This has driven £200m-plus in extra easyJet revenues annually.

As befitting a value-oriented brand, easyJet has achieved these results through a strategy of aggressive cost-cutting for its digital direct response marketing. The company has reduced its number of ad tech network partners from 25 to 3 and has also completely overhauled its search strategy. Instead of buying the one big brand term, they used more targeted buying. For example, the company stopped buying the search term "easyjet", then noted, through its Google Analytics, that site visits did not drop off as a result.

These days, if the company appears near the top of rankings for a particular route ("gatwick to malaga," say), no extra budget is spent on this term. And the marketing team keeps track of this performance in real time through simple Excel macros. "It just gives it a bit more rigour," Ian Cairns, the airline's head of brand communications, added. "We don't want to give money to Google if we don't have to."

This process has led to savings which run at around £8m a year for search alone, and the extra cash has been used in buying more broadcast advertising, including this TV spot from earlier in 2013.

easyJet's shift towards brand-building is grounded in econometric data undertaken with market research firm Brand Science, which revealed TV's effectiveness as a marketing channel. "People are seeing the ad and going straight to the website, with no need for buying a search term," Cairns added.

Until recently, CRM was also left relatively unexplored. "When I arrived at easyJet we didn't really have a customer database," Duffy said. "But now creating personalised messages is a big focus." The database itself now contains details for 50m people, with closer targeting achieved by easyJet bringing its data sets together: sources include customer booking details, contract history and social media interactions and mentions.

Mixing data sources in this way was cited as a marketing need by other presenters, who suggested that brand owners should not only rely on their proprietary data sets. Richard Marks, director of Research the Media, who has conducted research on this issue, written up in a recently-released report for UK agency trade body the IPA, made this point again and again in his presentation. "Customer acquisition requires access to the wider market," he advised. "Brands should not pull up the drawbridge - masters of their own domain, with a blurry view of what's outside. Big Data sets need to be slotted in to give the total market context."

Signs that media research companies are getting this message, and are making "mixed" data of their own available for brands to combine with their own proprietary sets, include Nielsen's public commitment to using return path data and, in the UK, BARB's new Project Dovetail, which offers the tantalising prospect of layering site-centric online data with traditional panel data. Justin Sampson, the organisation's CEO, also told the audience that it was set to collect audience data for TV streamed on tablets from 2014, thereby reflecting how increasing numbers of users watch shows.

And, despite Big Data's ubiquity as a phrase used at marketing conferences, Marks pointed out that there is a way to go before the industry gets its use of this data right. "If Big Data is the new Oil it still needs refining," he said. "Just because a dataset represents something in detail it does not mean it covers everything. Clarity is essential - knowing what you're working with and where it came from - and labelling is the key."

An expanded version of this post is available to subscribers in our Event Reports section.