Multi-Touch Attribution (MTA) can tell you a lot about which parts of your marketing are working, but does it tell you enough about what parts are not working?

multi-channel attribution in marketing

The origins of multi-touch attribution (MTA) were in the digital space, as a result of advertising spend transitioning away from traditional “offline ads” to digital media and channels which were deemed to be more accountable. Journeys within a client’s website, or between websites, could be stitched together and the resulting orders joined back to customers and their orders.

Now that the removal of third-party cookies is going to remove much of the stitching between websites, unless you are working with a collaborative data solution that allows this to take place, you are left just with customer journeys within your own website in full view.

These journeys may however include affiliates, referrers, digital ad campaigns, PPC, and direct search so you will at least know where the visitors came from if not the ad impressions they had been served to get them there.

However, we believe that there is still a great deal of merit in MTA, but not when it is restricted just to online events. (There is also the huge consideration that some of the large analytics platforms use sampled data and not 100% raw data which means the more you dig, the less you see).

Estimates vary about how much of advertising spend is digital, but the consensus appears to be around 55% currently, and that leaves 45% non-digital which is clearly far too much to ignore, much as Google would like us to. We also suspect that with the removal of some of the programmatic advertising volume, the digital proportion is likely to reduce down, perhaps to around 50%.

There is absolutely no reason why MTA should ignore the non-digital channels; but it means that you require the technology to joint it all together at a customer and order level. This is most effectively achieved using a customer data platform which is specifically designed to join browsing activity with off-line into a single customer view.

The non-digital ‘touches’, we prefer to call them ‘events’, can for instance include emails opened, text messages, call centre contacts, retail visits, and direct mail. These are all direct events, but on top of these are non-direct advertisements such as TV, which we discuss below.

There is a lot of unscientific opinionizing about the best approach to weighting events before an order. We are confident that we have found a reasonably good statistical solution for this. It uses a mix of Markov chains and survival curve statistics to give the weightings to any specific set of events. This approach does not presume anything about first or last touch, but rather looks at the evidence presented when the events have all been joined together in a single customer view, together with your customer and order data.

To deal with the non-direct channels like TV, the more ambitious will also want to build econometric models which reveal the overall effect on demand of all channels, direct and indirect, when working in combination. Econometric models often get a bad press as being unresponsive to short term changes in consumer behaviour, and not being granular enough in their spending recommendations, but they are the best tool we have to give the non-direct media their fair share of the credit for sales made.

Techniques now exist to align econometric models with multi-touch attribution so that, in effect, value initially credited to direct channels can be reattributed back to the indirect channels; this usually has a significant influence on the overall share of demand given to the direct channels.

So, to present what we have been describing diagrammatically, a full attribution process is going to look like this:

full multi-touch attribution process diagram

One of the often overlooked, and we believe very significant benefits of MTA, when it is sitting on a single customer view, is that it can be cut by customer type. The simplest cut is to distinguish between what is bringing new customers versus existing. But the cut can be for any customer segment, like high value versus low value customers, or purchasers of particular types of merchandise.

multi-touch attribution view of new vs existing customers

Multi-touch attribution can tell you a lot about how your marketing works, but only when you look at all of your online and offline channels in combination. And for many advertisers using indirect channels like TV, then it becomes important when possible to align MTA with econometrics.

In so far as we only look at events prior to a sale we will learn nothing about what doesn’t generate a positive outcome; however, if we take a look at all browsing events, we can start to examine the probability of an event leading, or not leading, to a sale.

There are two reasons why this is valuable information, although unfortunately often ignored. First, because knowing the probability of say a Facebook advert leading to a sale brings a sense of realism about advertising there, but also because serving people adverts in which they are not interested does damage to your brand.

Back in the heyday of direct mail, people were so fed up with the quantities that kept on arriving that they called it junk mail, and often had stickers on their post boxes asking for it not to be delivered. (Unfortunately, the postman had no choice but to pop it in their box).

PPC Protect estimates that in 2021 the average person (we assume in the US) will see up to 10,000 ads per day, whereas in 2007 estimates were only at 5,000 ads per day.

Common sense suggests that this must be way over the top of what is either necessary or enjoyable, and people will increasingly assert their objections to it.

Clearly brands that focus on the relevance of their advertisements will create a much more favourable impression than those that just focus on volume.

We have started to investigate browsing behaviour in terms of its likelihood to lead to a sale, with the following result:

Probability of browser moving to and from events and a sale
Probability of browser moving to and from events and a sale

To explain how this table works (and it was built using actual online and offline event data) it shows the probability of a person moving either from one event channel to another, or to a sale. So, if you start with picking a channel on the Y or vertical axis, you can then move along the row to view the probability of a customer moving to the next browsing state. For instance, someone coming to your website from a social network has a 96% probability of doing nothing further, and a 0.55% probability of being converted to a sale without engaging with additional channels. They also have a 0.67% chance of moving next to a search engine, whence they will have a 3.3% probability of making a purchase. However, someone receiving a campaign has a 6.5% probability of conversion without using other channels, and a 6.7% probability of moving next to a search engine.

So, in conclusion, we suggest that there is a strong role for multi-touch attribution, post third party cookies, with or without econometrics, and another new role for data science in investigating what we might call dark advertising, the stuff you see, but which makes little or a negative impression.

Read more about how Unifida’s marketing attribution works and what it can deliver.


UniFida logo

UniFida is the trading name of Marketing Planning Services Ltd, a London based technology and data science company set up in 2014. Our overall aim is to help organisations build more customer value at less marketing cost.

Our technology focus has been to develop UniFida. Data science business comes both from existing users of UniFida, and from clients looking to us to solve their more complex data related marketing questions.

Marketing is changing at an explosive speed. Our ambition is to help our clients stay empowered and ahead in this challenging environment.


Do you always have confidence and trust in your data to make important marketing decisions?

building trust in data to make marketing decisions

We are asking the question because we expect the key currency of the new post-COVID economy will be trust, and trust in data. 

Imagine you are amid your biggest campaign of the year, you are explaining the results to your leadership team, and are faced with questions like “how do you know”, “is the data right” or “why didn’t that campaign reach all of our intended audience”? I’m sure you have been in these types of scenario, which happen every day in the life of a marketer. 

SnapLogic [1] recently published an intriguing research report on how data distrust impacts analytics projects and decision making, which highlighted:

  • 77% of IT decision makers do not completely trust the data in their organisation for accurate, timely, business-critical decision making.
  • 76% of IT decision makers report that revenue opportunities have been missed due to a lack of data insights. 
  • 83% find data is not available at the time it is needed
  • 53% of mid-size companies suffer from too many disconnected data sources.

So, we would like to focus attention on some of the key data and insight issues faced by mid-size B-to-C companies in the UK and make suggestions around how they can be resolved.

Our experience is that these problems often have three separate causes:

  1. Customer data availability and quality
  2. Availability of skilled data analysts equipped with the right analytical tools
  3. A failure by the decision makers to frame the right questions for the analyst

 

Customer data availability and quality

The SnapLogic report reveals that 53% of mid-size companies have too many disconnected data sources, while 40% have poor integration of data sources meaning that data is missing or incomplete.

A typical B-to-C marketing department will often be looking at a distributed data situation with multiple silos like this:

data flow of different silos
Distributed data flow with multiple silos

The problem with this configuration is that there is no place for maintaining the overall customer picture, just pieces of the jigsaw in different places. So, it would be well-nigh impossible to answer questions like:

  • where am I acquiring my higher value customers from?
  • how is my latest email or catalogue campaign performing when most orders are placed without source codes via the website?
  • how do I know which of my dormant customers are worth trying to reactivate?
  • how many of my orders are coming from customers recruited this year, last year, and the years before?
  • how do I understand the ROI I am getting from each acquisition channel?

… and many more.

One solution to the data availability and quality problem is to introduce a customer data platform (CDP) that ingests data from all available online and offline sources and builds a single customer view. Marketers are increasingly focusing on first-party data to drive better customer experiences and marketing outcomes. More than half of marketers surveyed by Winterberry Group say cross-channel audience identification and matching is their highest priority. In fact, investment for identity resolution is projected to reach $2.6B in 2022, according to Forrester Consulting. So, it is no surprise that brands are taking this seriously and most want to create a single customer view.

A major part of what a CDP does is to undertake identity resolution; the process whereby data arriving from different sources is matched together using a range of different personal identifiers such as email, mobile, postal, cookie ID, customer number. The key consideration here is that the CDP needs to maintain for each customer a table of all known personal identifiers so that when a new one is introduced it can where possible be matched in.

The CDP then provides the single central source of truth about customer behaviour from which dashboards can run and analytics can be undertaken; it will also be used for activating multi-channel customer campaigns and for resolving GDPR questions.

 

Availability of skilled data analysts equipped with the right analytical tools

A large organisation like a bank will have upwards of 50 skilled data analysts, but with many smaller organisations it is often the case that they have one or none and rely on external resources to support them.

There are several reasons for this. Cost is a key factor and linked to that, the difficulty of putting a precise number on the value that a good data analyst can bring. Next the demand for analysis normally fluctuates, and a single analyst would always be facing feast or famine. Also, data analysts usually prefer to work in small teams so that they can discuss problems and learn off each other. Being the only data analyst in an organisation is a lonely position, and often they end up just cranking out reports and become dispirited.

A lot of the reporting can be resolved by introducing dashboarding technology like Tableau or Microsoft Power BI, but these tools still need to be configured to produce the right information.

However, dashboards and data visualisation tools can only take you so far. If you need some more complex analysis, or if for instance you want a propensity model to predict the next best offer to make to each of your customers, then a data analyst becomes essential.

To undertake more complex analysis the analyst will need good tools like SAS, SPSS, or R.

For the smaller organisations, the right solution could then be to outsource to an analysis company or to independent contractors, until demand has grown to a scale where the function can be brought inhouse.

 

A failure by the decision makers to frame the right questions for the analyst to answer

This issue is less frequently discussed but, in our opinion, not one to be brushed under the carpet.

A considerable amount of the work done by data analysts is wasted because someone does not spend sufficient time thinking about what the real problem is that the analyst should be trying to answer.

Einstein said…

“invention is not the product of logical thought, even though the final product is tied to a logical structure”.

Unravelling this statement in the context of customer marketing, we would suggest that the person who requests the analysis will succeed if they allow their imagination to fire up a range of conjectures that the logical analyst can then set about proving or disproving.

Some analysis is more mundane, but when for instance a business is contemplating several alternative strategic changes then the analyst should look at all the different scenarios that these would potentially deliver, and, as far as possible, provide the business with an understanding of their relative merits.

 

So, in conclusion…

From our experience it is fair to say that a large proportion, probably more than 50%, of medium size organisations involved in B-to-C marketing that we encounter have their customer data disconnected and spread across multiple systems. This is a problem that can be solved, and the costs are not frightening. A CDP will usually cost no more than 0.5% to 0.75% of sales.

However, setting up from scratch an internal insight and analytics department is challenging, and outsourcing will make economic sense until demand has grown. Also, the outsourced provider should have analysts with a very wide range of experience and skills.

And then how to ask the right questions of the analyst? We would recommend giving the analysts scope to try out different approaches, and to look at different angles to a question. Like this they are far more likely to land on an interesting and valuable solution.

[1] Data Distrust Report – the impact of data distrust on analytics projects and decision making published by SnapLogic in 2020, based on interviews with 300 US and 200 UK IT decision makers.


UniFida logo

UniFida is the trading name of Marketing Planning Services Ltd, a London based technology and data science company set up in 2014. Our overall aim is to help organisations build more customer value at less marketing cost.

Our technology focus has been to develop UniFida. Data science business comes both from existing users of UniFida, and from clients looking to us to solve their more complex data related marketing questions.

Marketing is changing at an explosive speed. Our ambition is to help our clients stay empowered and ahead in this challenging environment.


Exactly how much do online and offline channels contribute to sales?

Do you really know exactly how much each of your online and offline channels are contributing to your sales?

The blunt truth is that the great majority of marketers don’t!

Google for instance claims that businesses make an average of $2 for every $1 spent on Google Ads, but this is just a blatant case of Google marking its own homework. Are they saying that $1 Google Ads caused the $2 worth of sales, or just that there is some form of observed correlation? And how do they account for the impact of all the other forms of advertising from TV to outdoor to press to email?

Given the billions that hang on decisions about how to allocate marketing budgets across different online and offline channels, we felt that it was essential to work on giving our clients the tools and the knowledge to properly support these important decisions.

We started with some key assumptions:

– That decisions to purchase are necessarily complex and driven by multiple factors. Many of these factors like brand awareness cannot be recorded in the context of an individual sale, but many can be, and for those that are, we should look at all the known recordable events before a sale, and certainly not just online events, or worse still, just last clicks.

– What we call recordable events are activities like receipt of a catalogue, opening an email, visiting a website from a social media advert, or using Google Ads to find a website. Some of these events are driven by the customer like natural search, and some are driven by the vendor like receiving a catalogue.

– That we would assemble all the recordable online and offline events before each sale and use these as the dataset from which to analyse the true impact of different channels and different time intervals between an order and an event.

– That having overcome the challenge of collecting all the online and offline events together, we would focus our analysis on the central question of how to weigh the different types of events. Intuition tells us that an event 60 days before an order may have played a smaller role in the decision to purchase than one on the same day as the order, but the question we needed to answer was by how much? Also, should we give different weightings to different types of event? Is an opened email more or less important than a website visit happening as a result of a click through from a Google Ad?

– We do not want to suggest that marketers should ignore unrecordable events such as TV viewing or driving past an outdoor lightbox. Rather that their effects need to be analysed using different techniques like time series analysis, and that in so far as credit is given to recordable events it should be shared with the credit due to unrecordable events.

We have used orders and recorded events from two very different retailers to provide the data sets for our analysis. The first and surprising discovery was to find just how many recordable events actually happened. One reason for this is that customers may make multiple visits to a website before purchasing or open an email multiple times. The following table shows the number of recorded events that preceded each order in a 90-day time window:

the number of online and offline events resulting in sales orders

The analysis is ongoing, and we are aiming to publish a white paper on it in August, but there are three important initial findings that we can share:
– different channels should carry different overall weights, and we can analyse what they should be
– that each channel has its own time decay curve. In other words, the impact of events in one channel will wear off more quickly than for another channel.
– that the set of weightings used for new recruits should be different to those used for existing customers

The final results will include quantification of these findings.

If you would like us to share the white paper with you when it is ready please email us. It will be free for our newsletter recipients who order it in advance, but will be sold to others.

And if you would like meanwhile to have a chat to us about how to solve your own marketing mix attribution problem, we would be delighted to discuss, so please get in touch.

 


UniFida logo

UniFida is the trading name of Marketing Planning Services Ltd, a London based technology and data science company set up in 2014. Our overall aim is to help organisations build more customer value at less marketing cost.

Our technology focus has been to develop UniFida. Our data science business comes both from existing users of UniFida, and from clients looking to us to solve their more complex data related marketing questions.

Marketing is changing at an explosive speed, and our ambition is to help our clients stay empowered and ahead in this challenging environment.


Do you really know whether your company should install a customer data platform?

data from a customer data platform

Customer Data Platforms (CDPs) are taking the marketing world by storm – the Customer Data Platform Institute projects that marketers will spend $1.3 billion on them in 2020.

But the important question for your company is whether you would really benefit from having one?

So, what do CDPs really do? At a very high level they:
– ingest all available sources of online and offline customer data and build a deduplicated single customer view
– provide the capability to profile and segment customers
– enable personalised and consistent communications to take place across all channels by connecting your marketing technology
– support you in visualising customer performance

Do you really know whether your company should install a customer data platform? We have devised a simple 10-point questionnaire to help you understand whether your company could benefit from a CDP. It won’t be telling you whether should definitely should have one, as you will need a business case for that, but it will tell you whether CDPs are worth investigating.

QUESTIONNAIRE

If you want to start to find out more about what a CDP could do for your company then please order our free e-book CLICK


UniFida logo

UniFida is the trading name of Marketing Planning Services Ltd, a London based technology and data science company set up in 2014. Our overall aim is to help organisations build more customer value at less marketing cost.

Our technology focus has been to develop UniFida. Our data science business comes both from existing users of UniFida, and from clients looking to us to solve their more complex data related marketing questions.

Marketing is changing at an explosive speed, and our ambition is to help our clients stay empowered and ahead in this challenging environment.


Personalisation in digital marketing: how personal do you want to get?

There is a lot of interest amongst marketers in personalisation in digital marketing. And yes, we all like to be recognised for what we are. But we suspect that people are becoming wary of trivial personalisation or ‘personalisation lite’.

For example, recognising what I last looked at on a website is reminding me of the blindingly obvious.

However, telling me about a local retail event in the near future that is connected to what I was looking at on a website is much more likely to catch my attention.

And, for a full version of personalisation, understanding that I am a longstanding and loyal customer, rewarding me for my loyalty, and telling me about something that is actually of interest to acquire, really does hit the spot.

So, what is required to make this deeper level of personalisation actually work in practice for digital marketers?

Well, the first item needed is an effective personalisation engine. A tool that decides what kind of personal experience I should receive based on the information it knows about me. These engines work by the marketer setting rules or conditions which if satisfied mean that the customer will be sent a specific version of an email, or see a particular image, or offer on a website.

But next you will need to have joined together the online and offline data worlds. That allows the personalisation engine to know both that you are a loyal customer and that you are interested in a particular kind of merchandise or customer experience.

Most personalisation engines deliver ‘personalisation light’ because they ignore the history of your relationship with the company and just focus on what you have been browsing in the last 24 hours or so.

We have teamed up with Fresh Relevance (see www.freshrelevance.com) as our personalisation engine partner to allow you to give your customers full personalisation.

We manage this by sending Fresh Relevance a nightly feed of customers’ characteristics – factors like loyalty, spend, and long-term merchandise category purchases, so that these can be used in combination with their immediate online activities and behavioural triggers.

UniFida and Fresh Relevance can deliver in combination personalisation that really makes a difference.

To find out more, please email us suggesting a time that would be good for you to have a chat.


UniFida logo

UniFida is the trading name of Marketing Planning Services Ltd, a London based technology and data science company set up in 2014. Our overall aim is to help organisations build more customer value at less marketing cost.

Our technology focus has been to develop UniFida. Our data science business comes both from existing users of UniFida, and from clients looking to us to solve their more complex data related marketing questions.

Marketing is changing at an explosive speed, and our ambition is to help our clients stay empowered and ahead in this challenging environment.


Attaining a multi-touch attribution strategy

‘Attaining a multi-touch attribution strategy that works is like looking for the holy grail’. This is one of the conclusions in a report just published by the CallRail Research Unit (click here to download the report).

A key finding from their survey was that ‘36% of marketers say that lack of insights into the effectiveness of tactics, or an effective attribution capability, is the most damaging factor to their marketing efforts; a further 25% ranked it the second most significant factor’.

It so happens that we have recently completed developing a multi-touch attribution capability and it’s now part of UniFida.

Attaining a multi-touch attribution strategy enables you to understand the relative influence, and ROI, of all your online and offline marketing channels and media.

We would like to give you a live demo and show you how it works. If you can spare us 30 minutes, please send us an email and suggest a convenient time for you.

 


UniFida logo

UniFida is the trading name of Marketing Planning Services Ltd, a London based technology and data science company set up in 2014. Our overall aim is to help organisations build more customer value at less marketing cost.

Our technology focus has been to develop UniFida. Our data science business comes both from existing users of UniFida, and from clients looking to us to solve their more complex data related marketing questions.

Marketing is changing at an explosive speed, and our ambition is to help our clients stay empowered and ahead in this challenging environment.


What’s indispensable for your marketing data ecosystem?

What should be the ‘condicio sine qua non’ of all marketing data ecosystems?

The answer is very obvious, but very often overlooked; a marketing data hub or customer data platform that joins together all the front end and back end data.

Underpinning virtually all the multitude of martech applications there is the ever-present need for a solid data hub into which, and off which, they can all feed.

For instance, take website personalisation; it clearly doesn’t make sense to focus the nature and content of a tailored customer experience based just on their recent browsing, when you could also know if that individual was a loyal and steady customer or someone who was only cruising for sale offers.

Or how can you respond to a subject access request under GDPR if your email service provider and order processing system are not in some way linked around individual customer identities.

So what are the key elements in a customer data hub?

  • A means of joining together every data item that relates to an individual using all possible match-keys from cookie IDs to postal addresses
  • Persistent ingestion and storage from all on-line and off-line sources of every item of data from individual transactions to inbound or outbound contacts without summarisation or concatenation
  • Accessibility to other systems both for ingestion and exportation of data that is relevant for that application

There is also the critical organisational element; the data hub is far more likely to be successful, and to provide value, if it is owned and manged by marketeers who get the reason for having it.

This doesn’t mean that the development of a data hub is simple and not needing to be built in collaboration with IT experts. Data is also often untidy or in need of modification like miss-spelt addresses.

There is also a plethora of potential data sources to be fed into the hub, such as:

  • Order processing systems (e.g. for order read donation for charities, and policies for insurers)
  • Website browsers
  • Call center contacts
  • Email service providers
  • Third party data sources like lifestyle overlays or prospect lists
  • Identity resolution tables such as gone-aways
  • Loyalty card applications
  • Appointment booking systems

to name but a few we have encountered.

And then the hub once fed has to drive an ever-growing array of marketing tools. At a very high level these tools either support insights or actions. Typical examples include:

  • Dashboards and data visualisation
  • Digital personalisation
  • Email service providers
  • Campaign selections and response analysis
  • Contact centres
  • DMPs for digital media targetingCustomer research

Our view is that marketeers who plan their technology eco-systems without first planning a customer data hub are giving themselves an ever-growing problem. On the other hand, get the data hub right, and there is no limit to what can grow out of it.


UniFida logo

UniFida is the trading name of Marketing Planning Services Ltd, a London based technology and data science company set up in 2014. Our overall aim is to help organisations build more customer value at less marketing cost.

Our technology focus has been to develop UniFida. Our data science business comes both from existing users of UniFida, and from clients looking to us to solve their more complex data related marketing questions.

Marketing is changing at an explosive speed, and our ambition is to help our clients stay empowered and ahead in this challenging environment.