Your company may already be using attribution models to understand the impact of each moment in your customer’s journey – if so, you are one of the 76% of marketers who say they use or plan to use attribution measurements in the future. marketing Attribution measurements are important because they meet the customers demand for a relevant, fulfilling, multichannel experience.
Essentially, they enable you, the company, to understand your marketing down at the user level, and accurately assess the impact of each touch point or tactic you use in your marketing model. By learning which channels are doing best against each other, you can find out where best to spend on your marketing and adverts.
Also Read: Trends And Challenges 2020 For Digital Marketing Experts
However, there are mistakes made when measuring attribution – the model may be flawed, the data gathered may be poor quality, and even the way the data is implemented leveraged can be poorly done. It goes without saying that the more you know about the common mistakes and biases that lead to misattribution, the more intelligent choices you can make when devising a future marketing strategy.
Correctly done, attribution provides you with a window to view how your marketing efforts have succeeded or failed. Marketing in the modern era really relies on data – for retailers, this means the ability to follow the customer journey, from the moment of seeing a product to the moment of buying the said product. This might take the form of the customer opening a sales email promotion, clicking through the link to browse the online store, and later being retargeted with a relevant advert. The customer then eventually enters the local store and purchases the product they have been looking at online. An affective attribution model shows the influence of each touch point in the process, from the email to the advert to the physical store experience, and tracks brand impact and location. Effectively, attribution should measure both online and offline marketing channels and provide a complete view of a customer’s journey.
There are good ways to measure attribution, and bad ways to do so. Here are the most common mistakes made when measuring marketing attribution – and how to avoid them!
1: Isolating Offline and Online Attribution
Customers often engage with both online and offline marketing tactics as they move down the sales funnel – yet many marketers measure these sales tactics separately, rather than as two parts of one whole.
Refer to the previous example of a shopper first viewing an email marketing campaign, online adverts, and then going to the store to buy a product. It’s extremely important to see this is a cumulative process, not two isolated journeys.
Viewing offline and online marketing separately results in limited visibility across interactions a customer makes on their journey. It’s difficult to gain insights and improve marketing efforts if a unified view of customer engagements with marketing techniques is not available; if only measuring one or the other, huge parts of the process are left out.
In many cases, phone leads have been proven to be as valuable as online leads. Use call attribution to ensure all conversions are checked!
2: First or Last Click Models
Placing value on the first or last touchpoint in the sales process – that is, before the proto-customer is converted into a sale – is a poor attribution method. “The customer journey is often varied, fragmented, and influenced by several factors. As noted, before, an email and advert can lead to an instore purchase – yet using first or last touch attribution degrades the importance of everything else involved” says Chelsey Collins, a business writer at Writinity and Lastminutewriting.
Last-click attribution models are easy to handle and have low data requirements. But this a compromise for the limited and even skewed perspective they offer on the customer’s journey. Adobe reported recently that 79% of people switch devices in the middle of an online activity – so customers are frequently bouncing between many channels, and with last click attribution models, it’s impossible to see how all marketing modes work together. You might do something rash like decreasing your budget somewhere that is actually driving initial engagement with your products!
If you are a company only looking at first or last touchpoints in terms of driving conversions, you are missing out on a huge amount of data on what is truly influencing your customers.
3: Not Using Single Customer IDs
In the world of omni channels, it can be hard to keep track of who is who – but the easiest way to avoid duplicate leads is to have a single customer ID. This basically means even if a customer is using many channels, your attribution model will realize this, usually by using some for of cookie tracking for the online side. For the offline side, call intelligence is useful for phone leads. In short, ensure all your data is integrated, single customer IDs are in use and duplicate records are avoided!
4: Only Looking at Direct Response Campaigns
Direct response campaigns make attribution very simple – but marketers should still look at other types of touch points, like display impressions. Display impressions include upper-funnel media, branding and awareness campaigns which are often not thought of as revenue drivers yet are nonetheless part of the attribution model. Customers might not have clicked on a display advert, but the impression made by this advert still influences their purchase.
5: Not Measuring In-Campaign
In new marketing analytics platforms, which are fresh takes on old attribution methods, a crucial element is often missing – the ability to measure in-campaign. This basically means that attribution models focus on ‘lagging’ indicators, like the sales generated, remaining one step behind the target audience. It’s important to use attribution models that incorporate both leading and lagging indicators, enabling you to optimise your marketing model whilst it happens.
6: Using Cheap Inventory Bias
A common misattribution mistake comes from using cheap inventory bias. This means that low-priced retail goods that generate more customer engagement, it’s usually just because they are cheaper and so it’s more likely people can afford them. This simple answer is often overlooked by companies and marketing attribution models, attributing more value to the items and sales tactics themselves rather than just the price of the goods. This focus neglects the fact that higher priced goods are selling worse simply because they are more expensive, and its less likely customers are converted to sales through them!
7: Not Looking at the Lifetime Value of a Customer
When looking at leads, you should be looking beyond the initial conversion and onto the future.
“Not all leads are created equal – it’s important to realise this, and what the lifetime value of a customer might be. If a customer will repeat buy, purchase from you for a long time, or purchase goods often, they are a good lead. Good attribution models look at the average customer lifespan, customer retention rate, and the profit margin per customer. This allows you to see the lifetime value of customers and is incredibly useful for understanding what leads are going to be repeat-buyers,” notes Kathryn Knott, a marketing expert at Draftbeyond and Researchpapersuk.
8: Not Using Good Marketing Attribution Tools
To be fair, there are many good free attribution tracking solutions out there, but most of the free and lower-grade marketing attribution tools only work to a certain point. Every marketing technology stack should be suited to the business as each differs in the details – but in general, your CRM should track customer interactions through the entire lifestyle, use a marketing analytics solution to measure and analyse the path of a customer from first impressions to purchasing, and a data management platform to help you understand it all. Data is king!
9: Not Including all Working Parties
Attribution strategy usually involves a lot of different employees and teams, not just the marketing team! You need to align the entire company and all people involved, be aware of overall objectives from your CEO, and set your CIO to be aware of the needed data points. Your CFO sets the budget of your company and should be involved in understanding why the attribution strategy works, why it doesn’t, and to justify the spend on it. Of course, your CMO is the main manager and driver of the marketing attribution strategy. Attribution conversions shouldn’t only happen in the marketing room – everyone in the company should be involved!
10: Tracking Data that isn’t Important
Data is king – but not all data is useful. Attribution models are not exact and static, and it’s not an exact science, after all. It’s usually impossible to track the entire path to purchase of a customer with 100% accuracy. To achieve 100% accuracy, you’d probably have to break some data privacy laws in the process!
So, it’s important to understand what data from attribution you should be using. Decide what information has the biggest impact on future marketing strategies and decisions, and what will actually help you with optimising your marketing strategy. Then you can see what data you need to be collecting and searching for, and what you don’t.
11: Cookie-Based Measurements
Relying on cookies is detrimental to marketing attribution models for a couple reasons. Marketers relying on cookie-based measures to understand where and when a customer was influenced to purchase are missing out on data – because people can easily delete their cookies. When they do so, all this information is lost, leading to faulty representations of attribution and customer conversion rates.
To prevent this from happening, companies should focus on identity tracking rather than cookie-tracking to understand how their online presence is being used by customers.
12: Tag Capitalisation Mistakes
When using data analytics software like Google Analytics, there are a few mistakes in data management that can completely skew your results.
Sending data that is wrong to Google Analytics causes it to make flawed assumptions, and so data gleaned from Google Analytics cannot demonstrate how well a channel or marketing strategy is performing.
The first of these is tag capitalisation mistakes. Google Analytics tags are case sensitive, so ‘Nike’ is different to ‘nike’. This might seem small, but it can lead to incorrect assumptions about searches, site traffic, and lead to poor data segmentation.
It can be easily remedied by adopting a tagging culture throughout your company, ensuring everyone is using the same tags to reduce problems when reporting data.
13: Self-Referrals
You may have noticed with Google Analytics reports that your company’s own domain can be a referral to your site – which is a pretty chicken-and-the-egg scenario, as how can your domain be a referral source to itself?
It’s due to a flaw in Google Analytics, as GA ends a session automatically after 30 minutes of inactivity. A user will often leave a tab open to continue their session later, and this then messes up the analytics result of how that site was found. Whilst an organic Google search will show up on Google Analytics as ‘google/organic’, after a period of inactivity, that referral will be the site itself. This of course does not mean that the site referred itself – it was an organic google search.
You can fix this issue by adding your domain to the referral exclusion list, in the ‘Admin’ area of Google Analytics. This allows GA to attribute your traffic to the original source, rather than using your website.
14: Faulty Channel Groupings
The last attribution marketing mistake is to do with channel groupings – these provide a quick way associate all traffic with a certain source, to see how visitors from that channel interacted with your website. ‘Organic search’ is one such default channel grouping in Google Analytics, allowing you to see how users form a variety of search engines interacted with the site (Google, Bing, Yahoo, etc.)
The problem with this is that Google doesn’t recognize all mediums and categories, and so lumps these all into a channel grouping named ‘other’. You will need to go into this channel and check what sources are listed under other and try to recategorize them. By cleaning up your default channel groupings (again, via the Admin page), you can ensure your data is organised correctly.
Ashley Halsey is a professional writer at Luckyassignments.com and Gumessays.com who has written for a variety of papers and journals. Mother of two children, she enjoys traveling, reading and attending business training courses.