Econsultancy and IBM’s BusinessConnect series of events stopped off in Bangkok last week, giving local marketers the chance to share their experiences in data-driven marketing.

The Thailand conference was sandwiched between similar roundtable events hosted in Malaysia and Singapore.

The roundtables were a chance for delegates to discuss the main trends and challenges relating to data-driven marketing, personalisation and marketing performance management.

Here’s a summary of the main talking points, and to find out more about marketing in this region download our report on the State of Social Media in Thailand.

Where to begin?

One of the main themes emerging from the Bangkok roundtables was a sense that marketers almost felt slightly overwhelmed by the amount of data available to them.

Delegates were eager to step back from discussing very specific technical issues, and instead look at where to begin with their data analysis.

For instance, how do you work out which data is most important, and which insights are actionable?

How can we begin to make sense of it all and also start to use this data to make more money?

This is an issue that’s obviously made more difficult by the fact that few companies have the necessary infrastructure in place. All too often we heard people mentioning the problem of data silos. 

Data silos

For data-driven marketing and personalisation to become a reality companies need to have all their data in one place.

However very few, if any, of the delegates had successfully managed to merge all their customer data.

More commonly we heard that companies are struggling with siloed data and have no real strategy for bringing it all together.

In some cases, this meant that multichannel businesses could not give their in-store staff access to online purchase data, even for customers who held an account with the company.

Setting assumptions aside during data exploration

A number of delegates said they are concerned about making assumptions around customer data.

It can be difficult to approach data objectively and let the numbers speak for themselves, rather than imposing our own preconceptions or giving extra weight to data that supports our cause.

To avoid bias, it’s important to take a step back from the data in order to make sense of it.

Lack of maturity in data modelling

Delegates identified four different types of data that are important to marketers – demographic, transactional, interactional and behavioural.

A lot of attendees were only utilising the first two, which means they aren’t making the most of the opportunities that big data presents.

By failing to look at interactional and behavioural data marketers will be missing out on important customer insights and trends that could improve their campaigns.

This was a topic that came up on both the personalisation and data-driven marketing tables – it seems that Thai marketers are yet to realise the full potential of behavioural data.

Data collection and privacy

Data collection proved to be a popular topic – both in terms of what to collect and how it should be captured.

There can be a temptation to try and ask for too much data upfront, when a staggered approach might be more effective, i.e. just ask for an email address then request additional information at a later stage.

Asking users to fill in large data capture forms increases the abandonment rate and can also impact data quality, as people will enter false information just to get through the form.

Though marketers recognise the benefits of designing more concise data capture forms, there’s often pressure from elsewhere in the business to ask for more.

Incentives and freebies work very well in Thailand and can increase form completion rates, though it’s an expensive method of data capture. 

Mobile and local personalisation

Mobile and localised methods of personalisation were seen as hugely powerful for marketers, with discussions centring round the use of apps for content delivery.

For example, multichannel retailers can use apps to deliver product information or details of in-store promotions.

It could also be effective as a post-sales tool to improve the customer experience, so companies that previously only sold hardware can now also provide personalised digital services to deepen the relationship with customers.

This is a topic discussed in further detail in our new report, A Marketer’s Guide to Wearable Technology.

One important point to bear in mind is that personalised messages driven by mobile and localisation come with an increased risk of annoying customers or appearing invasive. 

Measurement and attribution

Delegates discussed their problems with measurement and how they can try to understand the whole customer path to purchase.

Which metrics should we be looking at? Sales and conversions are the ultimate goal, but what other factors are important throughout the customer journey to indicate success?

This ties into a bigger issue with attribution, and how marketers can justify investment in different channels.

In general it seemed that delegates from online businesses found the process of measurement a little bit easier than those from predominately offline businesses.

Importance of planning ahead

Problems with measurement and analysing the customer journey can be overcome to an extent by better planning.

If campaign objectives and the user journey are properly mapped out ahead of time, then it means we know what metrics to analyse in the reporting stage.

This can also allow marketers to more easily see which content and media channels had the greatest impact on conversion rates.