CXL Institute Growth Marketing Minidegree review Part 7

Picture from CXL course on Attribution
  1. Sort and Filter: The basics of making sense from data starts with sorting and filtering data. Sort can be done on ascending and descending manner whereas with filter we can reduce the view of the overall data to see the exact variables we want to.
  2. Sum variations: I never knew the different sum variations.
    a) SUMIF: sum based on one condition
    b) SUMIFS: sum based on multiple conditions with the use of relative and fixed values.
  3. Count Variations: Count variations were really cool to understand.
    a) Counta: It will return the values of text count (Add -1 if the text has header), count returns the number of count of numbers.
    b) Countif: Count based on one condition
    c) Countifs: count based on multiple conditions (similar to sumifs command)
  4. Table headers: We then talked about the general rule of keeping the headers containing numbers aligned to the right and the headers containing text aligned to the left. These rules can be changed in case of a filter is applied in which case the headers containing numbers can be aligned to centre. Furthermore, if we showcase the data into excel tables then any change or addition gets implemented throughout the table in the new formatting, this is an easy way to add formula throughout the table.
  5. Pivot Tables: I then got a refresher in pivot tables which can be created by data>pivot tables. We can then add rows and columns and filters to trim down our view. There are some really amazing features fro segmenting the data in the pivot table that I was unaware of:
  1. I learned about removing duplicates, splitting the data into text to columns and also that there are always some other way to achieve a result in excel.
  2. Vlookup is such a powerful formula that can be used in so many different situations. We looked at how to reduce the impact of those error messages through the use of IFNA or IFERROR. This command helps to avoid #N/A returns and rather return a controlled message.
    =VLOOKUP(search_key, range, index, [is_sorted])

Understanding string (text) functions

  1. LEN: Throws the length of the text
  2. SUBSTITUTE: Throws a substitute value for the part defined
  3. FIND and SEARCH: The find is case sensitive and search is case insensitive. So you can search what you are looking for, within where and how many letters in you wish to start the search.
  4. MID: Pull out some text some place in whatever string you are looking for
  5. LEFT and RIGHT: Pulls leftmost and rightmost characters from the string

Understanding Error Trapping

  1. Data validation: Reducing error and reporting error is a great feature of excel. This can be done through Data Validation.
  2. Pull down list
  3. Whole number specification
  4. Summing with brackets
  5. IF function to reduce the cross footing
  6. Protected cells to protect the sheet and keep a designated data entering space

Attribution

  1. Last non direct click: This is something that Google analytics use. This model gives the credit to the source where the second last interaction happened and not the last direct click.
  2. Last interaction: This gives the attribution 100% to the last interaction. This is very useful to measure online-offline conversion.

We then moved to tactical understanding against every channel

  1. CRO: Conversion rate optimisation typically focuses on improving the conversion rate on a singular interaction. Attribution in CRO focusses on understanding not just the conversion from that singular interaction but the overall interaction leading to the group of people converting.
  2. PPC: Attribution is powerful in PPC. Generally, the Google Ads attribution is based on last click or last non direct click models. Any change in the attribution model will impact the google ads account. We also need to understand that brand ppc is different than the generic ppc when we track attribution. While the former has short lead time to conversion the latter has a higher time and touch points. Therefore, to attribute the value the interactions need to be weighted accordingly.
  3. SEO: From an attribution perspective, we look at SEO from a customer journey. We can understand the terminology or the mood of the user looking at the search terms when they landed on a particular page. SEO is a broad channel, multitude of different search terms and every term will have different attribution. We need to map these keywords to the landing pages and understanding which landing pages link to which stage in the consumer journey.
  4. Display: Display works differently than any other source. There is a lot of challenge in determining the value to the source and getting the correct attribution status. The model that works in display is the post attribution model rather than the click based model. We also need to understand the campaign, the impression and the time of that impression. We can then map the user level impression to the user level journeys and attribute the impression before a visit and conversion. With the information on clicks and impressions we can then attribute the value of display.
  5. Affiliates: Attribution is really important for affiliates. Affiliates do take all the credit for the conversion but as marketers, I learnt that we need to have a better understanding of the customer journey, type, behaviour and understanding of the affiliate within the journey. With multiple affiliates out there, we need to map the customer journey and map the value of every interaction within the journey which might be at the awareness stage or the final conversion stage.
  6. Email: Understand the consumer to determine the emails to be sent to a user. There are multiple things that can be done to relay the information of an action within the email to the attribution model.
  7. TV: TV data is not in individual form but in aggregate. The TV attribution can be related with spot timing and the impact a tv campaign can have on SEO, PPC, Content and Email during that TV campaign. Attribute also helps us to understand the value of a such a campaign not just in that moment but till a few days, weeks after that campaign.
  8. Direct Mail: Every address of the customer should be assigned to a customer ID and all customer ID’s are linked to an attribution model. We need to load that piece of direct mail has been sent to the customer ID and if there was a smart URL that can be tracked with that customer ID. We need attribution for direct mail as it has the power to change the customer understanding of the brand and also the search terms.

Understanding the strategic shift due to attribution

  1. Customer journey analysis — We need to look at the consumer journey from the point of introducing the next best step. This will only happen once we understand the consumer journey and compare the conversion with yet to be converted consumers. Attribution will be able to showcase which action is more valuable.
  2. Value of Brand — There are 4 sections to this: Brand awareness and propensity, Product awareness and propensity. Brand awareness is for new brands, propensity is the inclination of an individual to buy your brand. In the product realm, we are trying to push that our product is better than the other products out there.
  3. Life time values — LTV is not an average. We want to understand the value of the consumer throughout the time he/she has been engaged to our product. The basic concept here is to understand that the cost to acquire a new customer is always high (10x) high then to re-acquire someone. Furthermore, the value brought in by the reacquired customer is 10% higher.
  4. Customer Data — Though we look at an aggregate value in attribution but the best part about this is that we can drill down to an individual level. This allows us to look at some of the outliers from the usual customer journey that can even help in fraud detection or realigning our efforts to get those outliers to convert.
  5. Value of the user — This can be useful to lead scoring and propensity scoring. On every individual score the attribution model can pass a propensity score to that user. This can be used by multiple tools like bid management, CRO tools or lead scoring systems. We can then overlay this to the people who have not converted. We can then understand the propensity to convert with all the users who will not convert and tailor your outreach and marketing activities accordingly. So attribution at an individual value of the user is really important and valuable.

Thoughts:

Startup Growth | Business Development and Sales | Creative thinker

Love podcasts or audiobooks? Learn on the go with our new app.

Recommended from Medium

Data Pre-processing tasks using python

Internal Nasal Dilator Market Size, Share, Growth, Segment, Trends, Developing Technologies…

5 ways to get the most out of mWater

Global Optic Nerve Disorders Treatment Market Analytical Overview, COVID-19 Impact Analysis…

Science Doesn’t Stop when the Art Starts: 9 Steps to Equity & Ethics in Data Communication

Can Data Science Help Us Save Animals?

🦉 10x curiosity — Issue #154 — Bayes — How Can you be less wrong? -[ FROM THE VAULT]

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store
Alakshyendra

Alakshyendra

Startup Growth | Business Development and Sales | Creative thinker

More from Medium

A Complete Guide To B2B Lead Generation For Sales Professional |Ampliz

B2B Lead Generation For Sales Professional

THE CUSTOMER ACQUISITION PLAYBOOK

Five Things You Can Do to Unlock Empathy in Your Marketing Today

Google Analytics 4 Is Coming — What Does This Mean For Retailers?

Image of Google Analytics logo with physical retail image