Abstract— The ASAP Playbook aims to provide personalized and actionable insights to the Account Team, primarily Account Executives and Sales Reps. It empowers them by informing them what accounts to focus on, and which products to pitch with the right ‘Call to Action’.
1. INTRODUCTION
As part of FY20 Client Solutions Group (CSG), there were a lot of workstreams and projects going on within the team. Few of the key ones:
- Enabling the transition to Win 10: Insights and recommendation for Accounts to move to Win 10 as Win 7 was going End of Support (EOS) in Jan 2020
- Strengthening Market position in Core Notebook Category: Providing recommendations and insights about most likely new generation buyers and Latitude Series 7 to series 5. Migration for accounts buying 7- series on low margin low price
- Widening Product Portfolio through Premium Products: Precision/Rugged Product recommendations along with Memory Acceleration Chip recommendations
- Expanding customer base through new buyers: Personalized recommendation for New/Non-Buyers based on Competitor Information and Seeds/Visit Information
There was a need for unified communication for multiple workstreams and key CSG programs. ASAP Playbook acts as a one-stop shop and a unified delivery engine providing personalized and actionable insights to Sales Reps.
2. METHODOLOGY / ANALYTICAL APPROACH
ASAP Playbook incorporates a variety of information coming from different data sources. The data sources can be classified into 3 buckets.
A. Databases:
• Warehouse – For historical transaction data
• Seed/Try & Buy Data
• Engagement and Touchpoint Data
• Client Peripherals and Displays (CPnD) Recommendation Data
• CRM data – Expiring Tags
B. Marketing: Marketing teams provide the Campaign and Lead information
C. Insights and Recommendations from Individual Workstreams:
• Transition Analysis: This is a stacked model of Neural Network and Random Forest. The model helps to identify the accounts that are most likely to buy new generation products.
• Win10 transition: Provides a list of accounts that do not have a 100% Win10 mix and since Win7 was approaching EOS in Jan 2020 these accounts had to be transitioned to Win10. Win10 transition analysis is based on 5i-Bucketization.
• Another focus area was to help accounts in growing Direct Mix and protecting accounts that have high Direct Mix. This analysis is based on Time Series Forecasting.
• Memory Acceleration Solution: Chip recommendations are provided at a Quote Level so that a customer gets better configuration and enhanced performance at an affordable price.
After all this information is gathered, it gets plugged into a Multi-Criteria Decision Making/Analysis (MCDM/MCDA) which in our case is an 18-Factor Weightage Model to prioritize Accounts at an account executive Level.
What is MCDM?
Multi-criteria decision making (MCDM) is an algorithm that deals with decisions involving the choice of the best alternative from several potential candidates, subject to several criteria or attributes that may be concrete or vague.
When making decisions, decision-makers always try to choose the optimal solution. Unfortunately, an optimal solution exists only in the case of one single criterion; in real decision situations, almost any decision involves some conflicts or dissatisfaction.
How does MCDM work?
The process of MCDM involves the comparison of two or more alternatives. Each alternative is evaluated against multiple or n criteria according to the priorities of the decision maker. The degree to which an alternative satisfies a criterion corresponds to a utility value. The scores must then be combined to give an overall rating on that alternative. For instance, when a customer is planning to buy a new car, with so many possibilities he must evaluate every offer against n criteria, such as price, efficiency, service, and warranty conditions, to identify the most suitable option. The purpose of aggregation functions is to combine multiple numerical inputs, usually interpreted as degrees of membership, the strength of evidence, etc., into a single numerical value, which in some sense represents all the inputs.
Similarly, for ASAP Playbook we have 18 factors that go into MCDM, and ultimately help with choosing top accounts for the Sales reps to focus on.
The factors are classified into 3 buckets:
- Account Stage: It involves factors like Open Pipeline, Marketing Plays, Premier accounts, and Accounts expected to Grow Direct per our analysis
- Account Timing: Here the factors like Seasonality, inputs from Transition Analysis, Engagement, and Touchpoint Data are considered
- Dollar Value: Estimation of Client Potential
As an output of this exercise, we get a list of accounts ranked at an Account Executive (AE) Level. To simplify the whole process, let’s say an Account executive has 50 accounts under him/her. Based on multiple factors, the model ranks all the 50 accounts with a numerical score. These are not just the top accounts with the highest revenue/margin numbers but also the accounts which are best positioned to grow in terms of our priorities.
At beginning of every quarter, the top 16 accounts are listed in emails to the AE and the team, while the remaining accounts are kept with Marketing managers on the field as ad-hoc ASAP Playbooks so the AE can refer to any account under their portfolio if needed.
The delivery happens at the first 4 weeks of each quarter. In the first week, the ASAP goes out at the AE level containing the TOP 4 focus accounts for each AE. This way we cover a total of 16 accounts at an AE level which henceforth will be known as the TOP 16 accounts. The portfolio view is nothing but the summarized details of all the accounts under that AE. It helps them to have a very high-level view of how they are performing overall. It contains metrics like:
• CSG (Client Solution Group) share of Wallet: It represents how much Actual Revenue is generated from all the accounts under an AE by the aggregated Buying Power of all those accounts.
where:
x = A particular AE
i = number of Accounts under each AE
• CPD (Client Peripherals and Display) pen rate: This represents the CPD Revenue generated by all the accounts under a particular AE to the combined system revenue of all those accounts.
where:
x = A particular AE
i = number of Accounts under each AE
- CSG (Client Solution Group) Buying Power: This is the summarized view of the total Buying Power of all the accounts belonging to that AE.
The next part contains the client/CPD Revenue for the current and last FY along with the margin% of the current FY. Again, all these are rolled up at the AE level. There is also another interesting metric here, the Potential in both Client and CPD.
The Potential calculation is done using an Auto Clustering Method on Total Revenue per Unit (TRU), and Total Margin per Unit (TMU) at a Region > Industry > Quarters > LOB level. Once the clusters are identified, the Cluster TRU is calculated and is in turn used to calculate the Potential revenue for an account. A similar exercise is done for CPnD and then rolled up at an AE level to calculate the portfolio level potential.
The snapshot part of the portfolio view tells which 4 accounts would be recommended for that week and what we are recommending for them.
Figure 2. Account Level Overview
Next, for each of the Accounts which we recommend, the Account historical is displayed (Figure 2). The header contains the account name Account ID and Account Bands (High/Mid/Low/No). It contains the below sections:
A. In the “What they buy the most” section we show the highest bought LOBs (Line of Business) in the Last 7 Quarter + Current Quarter along with the TOP 3 Brands Per LOB.
B. In the next section the SOW (Share of Wallet), Client/CPD Revenue, Client Potential along with the LOB depth for that account.
C. In the “When They Buy” section the buying pattern of the account for the last 12 quarters is displayed.
D. Depending on the band in which an account is in, the next sections are displayed. If an account belongs to the Low/No Buyer band, we show them their competitor and Seed/Try Buy Data to have better chances and converting that account into a regular Buyer. But accounts in the High/Mid bands do not need such data, so for these, we show them instead their route to market and what they buy most in CPD
What we then have is the Call to Action (CTAs). Suppose an Account has got both Non-Buyer and Default Workforce Transformation then only Non-Buyer CTA will be displayed in Recommendation Box 1 as it has got a higher priority than Workforce Transformation.
Next, we will go through the CTAs one by one.
1. Generation Transition:
Desc:
In the Generation Transition CTA, accounts that tend to buy new products soon after the launch at a good margin are identified by using a stacked model of neural net and Random Forest.
In the CTA, the Prior Gen product, New Gen Product along with the historical customer behavior is shown.
Condition:
Any Account which is recommended by the model will have this CTA.
2. Non-Buyer POC (Point of Contact):
Desc:
This CTA deals with those accounts for whom we have very little info. This CTA describes where in CRM they can find contacts and which of the CRM’s Active contacts tend to play a part in the purchase decision.
This helps in giving the tentative POCs for an account on which we do not have many details.
Condition:
Accounts belonging to the LOW/NO Bands and which do not have a Generation Transition CTA will be shown this CTA.
3. Workforce Transformation:
Desc:
This is the default CTA of Recommendation. This shows the many ways in which Dell Products can help an Account with its Workforce Transformation.
Condition:
If the account does not have generation transition and does not fall in the LOW/NO bands then this CTA will be displayed.
4. Premier Accounts:
Desc:
There are 2 types of Premier Account CTAs. The first deals with those accounts in the High/Mid bands which do not have a premier profile setup. The second one deals with those that have a Premier profile setup but do not actively use those.
Condition:
If an account belongs to the High/Mid band but does not have a premier profile setup, it will get the Setup CTA, or else it will get the Premier profile Active CTA.
5. Core Notebooks:
Desc:
This shows the many ways in which Dell Products can help an Account with its Workforce Transformation.
6. Win 10 Transition:
Desc:
This CTA deals with Win 7 EOL.
Condition:
Any account which is yet to transition completely to Win 10 and has an open pipeline will get this CTA
7. Grow Direct:
Desc:
This CTA talks about growing or protecting the direct share of an account by leveraging additional discounts on specific products.
Condition:
Accounts with declining or low Direct Shares are identified using a time series analysis and sent to this CTA.
8. Cross-sell LFD:
Desc:
In this CTA we ask the sales reps to attach more Large Format Displays to the account.
Condition:
If the account does not have Win10 or Grow Direct CTA, then we display this CTA
9. Expiring Tags:
Desc:
This CTA deals with Systems that are at the end of the warranty period. It is the lowest-hanging fruit for a sales rep.
Condition:
Any account which has expiring tags in the Current and Next Quarter will be sent this CTA
10. Premium Products:
Desc:
Accounts in Premium Product play will get this CTA related to selling Premium Products like Precision and Rugged to those accounts.
Condition:
Accounts in the premium product marketing play and those who do not have Expiring Tags will get this CTA.
11. Cross-Sell LoBs:
Desc:
This CTA is complementary to the LoB depth section displayed in the account history. If an account does not buy a specific set of LoBs which they potentially can, they will be served this Cross-Sell LoB
The next section contains the CPD Recommendations. Here we mention what can be attached with brands they are already buying or, those we are recommending the sales rep to pitch to the accounts. This helps us provide an end-to-end Full Solution for the account starting from LoBs till attach.
3. DELIVERY ENGINE
Any report however fancy is not worth it if the target audience cannot use it. For ASAP this is taken care of by a custom-built, one-touch delivery engine. We have developed an R script that generates the ASAP Report for all the concerned AE from the reporting template. Next, we have a Python script that converts the excel into respective account PDFs, builds out the mail, and then delivers them to the concerned team.
4. DELIVERY MAIL
The main body contains the AE portfolio view along with details about the accounts, links/resources for additional resources, the ASAP attachments, and other additional attachments like Account Plan, draft emails, and more.
In addition to these, we also provide a survey link so that we can have feedback on what the AEs are feeling about the ASAP mailers. The entire account team of all 4 accounts is looped in the mails.
5. CONCLUSION
ASAP serves as a unified communication for key initiatives which comprises tailor-made recommendations at the Account Level. The insight generation is done through various statistical techniques and machine learning algorithms. The ASAP Playbook equips Account Executives with actionable recommendations and assists them to have better customer communication.
