Marketing Attribution: Covering Your Bets

0
72

Share on LinkedIn

Source: Pexels

Marketing attribution: An Overview

At casinos the house always wins in the long run. The odds are stacked in their favor but if you play blackjack, and count cards well, you can flip the odds on the casino. Marketing attribution is core to the measurement and optimization of marketing tactics and programs. When attribution is setup and used properly organizations may use the results to ensure their spend (bets) are covered and the return on investment is positive.

Marketing attribution is the process of allocating revenue, and customer & prospect behaviors, to specific sales and marketing interactions. A proportion of a customer’s revenue is attributed across all eligible marketing programs, that touched the individual customer, based on the attribution rules that are put in place. We will cover three facets of the attribution journey:

  • Attribution capability considerations and fine tuning your attribution process
  • Build versus Buy
  • Using the attribution output in ongoing marketing measurement and optimization

Attribution Capability Considerations

Marketing and sales attribution is a technical and analytic process. Multiple channels are involved including digital platforms which require full seamless integration. Given the technical nature of attribution, and all the nuances of your individual business and marketing programs that need to be considered, the following attribution capabilities should be assessed:

Attribution Touch Points: These are the campaigns and tactics that need to be included in the attribution process, based on your environment, which consist of marketing programs, sales interactions & events, and digital engagement.
Attribution Grain / Level: The individual and organization levels attributed including anonymous users, known prospects (or leads), individual customers and organizations.
Attribution Methodology: This is the underlying logic which attributes customer interactions and campaigns to resulting revenue. Assess the following based on your needs:

  • One-Touch or Multi-touch: Multiple customer interactions and marketing campaigns could be attributed to the same resulting revenue and if that is the case in your environment you may need a multi-touch solution where all eligible interactions can be attributed to the same revenue and the appropriate allocation is applied to each interaction.
  • Weighting Process: The attribution weighting by tactic type (paid digital media, paid social, email, event, blog, webinar,..) needs to be considered and a custom weighting capability may be required.
  • Modeling Techniques: As multiple interactions are attributed to revenue the cadence and timing of those interactions should be considered and there are multiple modeling techniques available to assess based on your needs. The primary techniques include the following in order or complexity:
    1. Linear where all interactions are assigned the same weight
    2. Ramp-up or time decay where increased weight is assigned to more recent interactions.
    3. U-shaped where more weight is placed on the initial and last interactions which drove revenue
    4. W-shaped: W-shaped models take into account both the timing of an interaction and the potential weight of each interaction category or tactic. The implementation of W-Shaped models can be both through business rules or a more complex data science models.
  • Attribution Results: In many environments revenue is attributed to each interaction but forecasted revenue may need to be considered as well as other prospect and customer behaviors.
  • Integration Channels & Solutions: Several integration channels and systems need to be considered. These include digital paid media (Google, Facebook, Twitter, Programmatic,..), CRM Systems, Marketing Automation Systems and Web Behavioral systems.
    Data Access & Technology: Having access to the attribution output data is critical in order to measure and optimize over time so consider if the solution is cloud-based, has a data API, and integrates with your CRM system. Data security will need to be assessed as well if it is a cloud-based system as in most instances sensitive Personal Identifiable (PII) attributes will need to be sourced.

    Once you have identified and rank-ordered your key attribution functionality, decided on and implemented a solution, the system should be continually evaluated. Many solutions have the capability to adjust weights by marketing tactic and some even have a data-driven weighting process as several factors can change the weights over time. Regardless of the weighting capabilities of your attribution solution revisit the weighting and modeling process every 3 to 6 months.

    Build versus Buy

    The process of evaluating the best long term option should include an assessment of available attribution solutions compared to internal development. The reality is that most will choose to purchase a solution but the build versus buy considerations include:

    • Technical Environment, Skills & Bandwidth: Is your technical environment ready for a custom attribution solution which requires seamless integration with paid media systems, internal data, web analytics systems (or the website directly) and CRM systems? Are the technical, business intelligence and analytical skills available in-house for internal development and will bandwidth be available for those resources?
    • Required Capabilities & Development: Map your needs to the capabilities laid out in the section above and estimate the cost of internal development for both the initial development and ongoing maintenance and enhancements.
    • Cost: Evaluate attribution solutions on the market that fit your needs and compare the cost to internal development and maintenance.

    If attribution is core to your acquisition efforts you will be well served to evaluate potential solutions and capabilities of those solutions. Even if you decide on an in-house build the understanding of the capabilities within fully evolved systems will help in your internal development.

    In general, if your environment is not highly custom, and the cost of internal development and maintenance is similar to purchasing an evolved attribution solution it could be best to consider a purchase, however be sure to carry out additional research and ensure that your final solution choice has a development roadmap which aligns to your current and future needs.

    Embedding Attribution in Measurement and Optimization

    Many of the available attribution solutions provide nice reporting interfaces which show the performance and value of the marketing campaign and tactics, Revenue may be one of the measures incorporated and if marketing spend / costs are readily available they may be included to calculate ROI. This solution-based reporting, or similar reporting you choose to develop internally, provides a means of understanding which campaigns and campaign categories (tactics) are performing well now and enables short-term marketing decisions that increase performance however the attribution data can provide and enable much more:

    • Media Impact (Media Mix) Insights: Attribution solutions tend to capture the campaign and interaction data well and better enable the development and application of media impact analyses and models. Media impact models look at the results of marketing interactions over a longer time period (preferably 1 to 2 years) and can measure the impact of branding as certain individual campaigns or channels may not provide a short term ROI but they can provide an increased longer term ROI through awareness and branding, Media impact models also help gauge the optimal mix of tactic types over time and allows for superior budget allocation.
    • Prospect Acquisition Models: Prospect or leads are tied to desired outcomes (revenue or other behaviors) within the attribution process and the resulting data is used to create robust acquisition models that may be applied in internal acquisition efforts and other external programs and channels.
    • Customer Retention and Up-Sell Opportunities: The attribution solutions track customer behaviors, and link those behaviors to resulting revenue, which may be used to create customer retention and up-sell models. The models may then be used to increase the performance of retention and cross-sell programs and strategies.

    Final Words

    Implementing and maintaining an attribution solution requires the marketing, technical and analytic teams to work together. It is important that every constituent understands the system & methodology well as making the process too complex initially could make it difficult to apply the attribution results – walk then run. Enhance over time as all three teams move forward together.

    Organizations do not like to refer to marketing and sales investments as gambling but it fits the definition of gambling if the efforts are not tracked, measured and optimized. In many environments attribution is a critical componant of the ongoing optimization process. Casinos frown on card counters, marketers do not!

    Roman Lenzen
    Roman Lenzen, Partner and Chief Data Scientist at Optumine, has delivered value added analytical processes to several industries for 20+ years. His significant analytical, technical, and business process experience provides a unique perspective on improving process efficiency and customer profitability. Roman was previously VP of Analytics at Quaero and Director of Analytics at Merkle. Roman's education includes a Bachelor of Science degree in Mathematics from Marquette University and Masters of Science in Statistics from DePaul University.

    ADD YOUR COMMENT

    Please use comments to add value to the discussion. Maximum one link to an educational blog post or article. We will NOT PUBLISH brief comments like "good post," comments that mainly promote links, or comments with links to companies, products, or services.

    Please enter your comment!
    Please enter your name here