If you've read through my other posts, you'll have seen that I've automated and optimised many processes on Adobe Campaign. The accumulation of improving our procedures eventually led me towards automating my first weekly CRM campaign:
In this post I will go through:
The previous execution process of daily CRM sends
Problems with the execution procedure and the goals I set out to achieve
Automating the data approval
Automating the click through data logic
How I personalised data using Adobe Campaign
The new campaign execution procedure
The benefits that automating the campaign provided
1. Existing Execution Process:
In Adobe Campaign, the CRM Data team had to execute workflows to send out emails to a specific set of customers who were targeted with this promotion. For daily gaming sends, it required me to do the following:
Check the targeting criteria of the workflow to make sure it matches what was requested by commercial stakeholders
Once done, I would then have to tick off the quality control checklist to pass auditing requirements. Each campaign has a checklist that would require sign off.
After ticking off the list, I would then need to manually execute the workflow, which would also extract the relevant set of customers to be credited with the promotion. (This process became quicker after creating the SNAP)
Once executed, the data would then arrive to an approval which would send me an email report of the waterfall breakdown of the data.
Adobe Approvals for Data - This approval (Tick Icon) would send out the email reports into our inbox on the waterfall breakdown of a campaign.The email report will then have to be forwarded on to the relevant stakeholders on outlook manually and approved from a data side as shown below.
Email report of a daily campaign showing the campaign name added along with the stakeholders relevant to the campaign. After forwarding on the report, I would then need to click on the "click here to reply ..." link to send me to an approval site.
Once the data side was approved, the relevant marketing team would then need to create the email and manually attach it to workflow.
Data has now been approved to 2nd approval where the Gaming team then place the correct template in the flowThey would then need to await further instruction from the Promo Operations (Promo Ops) team to check whether the customers have been credited.
Once sign off is approved from Promo Ops, the marketing team can then attach the relevant template and approve the campaign.
Marketing team have now approved their side and data is sent to a scheduler.2.Problems + Goals:
This process was incredibly manual and time consuming, leaving the CRM teams with little time to work on more projects valuable to the business. By automating this campaign, I hoped to provide massive time efficiencies and create a template of a new ways of working in sending daily campaigns.
Like the SNAP project, I broke down each stage of the process and tackled them as a single issue. This approach would lead me to automating this campaign and removing manual tasks for other daily sends which can be shown below:
Process Automation
When I first looked into automating gaming sends, I wanted to find a campaign with minimal changes in targeting. This way, the Data team wouldn't need to sense check the data every week, removing the need to tick off a checklist.
The campaign I decided to automate was the perfect fit as the targeting remained constant, the email creatives were relatively easy to automate in Movable Ink.
However, to fully automate the data side I needed to do the following:
Automate the data approval
Create click through data logic in the workflow
3.Automating the Data Approval:
For the data team, executing the data approvals and sending the email report to the relevant stakeholders was our biggest constraint. Historically, the team were using data approvals (Purple Ticks in Adobe Campaign) to provide this function, but after a little investigation, I found out a better way to do this.
In Adobe Campaign, there was a Node called an Alert. This used the same function in providing an email report but it didn't require someone's approval for the data to move through to the rest of the flow. This solved the first part of the process in removing the manual approval step.
However, using alerts still didn't solve how to send over the email reports to the relevant Commercial and Promo Operations stakeholders. Since stakeholders outside of CRM didn't have access to Adobe Campaign, we didn't know how to send reports directly to them from the system. Whilst I contemplated the idea of setting up email rules in Outlook to resolve this, I knew it wasn't a long term sustainable solution, given the large number of rules that needed to be created.
That's when I found out the assignee group function in Adobe Campaign and how to vary permission access for individuals that don't use the system. Creating an assignee group allowed me to add email addresses of all relevant stakeholders of a CRM campaign to one group. This enabled me to send email reports straight to my colleagues without manual intervention.
After creating these assignee groups, I modified the approval process into the diagram shown below:
Automated Approvals in a workflow to provide campaign report without approval needed
4.Automating the Click-Through Data Logic:
Afterwards, I looked at automating the click-through hyperlink of the email towards the promotion hub.
One of the bottlenecks that the Promo Ops team had was creating a different click through link every week for the same promotion. For this promotion specifically, it was different because of the date, as shown below:
April 9th 2019 : XXXXXXX_090419_XXXXXX
April 16th 2019: XXXXXXX_160419_XXXXXX
May 23rd 2019: XXXXXXX_230519_XXXXXX
(the Xs stand for the other part of the click-through link which always remained the same)
As seen above, this date followed the numerical format of DDMMYY and I needed to find a way for the date to be populated for the hyperlink automatically.
That's when I wrote an IF statement in Adobe Campaign to creating a data field that would populate this information, which can be seen below:
(Iif (toInteger(Day(GetDate())) > 9, toString(Day(GetDate())), toString("0") + toString(Day(GetDate()))))+''
+(Iif (toInteger(Month(GetDate())) > 9, toString(Month(GetDate())), toString("0") + toString(Month(GetDate()))))+''
+Substring((ToString(Year(GetDate() ))),3 ,2 )
The yellow day formula can be interpreted as follows:
If the day of today's date is greater than 9, then convert that number from an integer to a string format.
Else, if the day of today's date is 9 or less, then convert this number from an integer to a string and concatenate it by adding a zero at the front. E.g for the 8th it will be 08 instead of 8.
The purple month formula can be interpreted as the same as the day formula but we substitute it with month instead. E.g:
If the month of today's date is greater than 9 (i.e October/ November/ December which is numerically known as 10/ 11/ 12), then convert that number from an integer to a string format.
Else, if the day of today's date is 9 or less (i.e. January to September), then convert this number from an integer to a string and concatenate it by adding a zero at the front. E.g for August it will be 08 instead of 8.
The green year formula brings the current year of the current date and converts the last two numbers into a string format.
Once this was done, all I needed to do was implement it in the workflow and the date would be shown in the relevant format as seen below when:
5. Personalising Data in Adobe Campaign:
After, creating the click-through data field, I was ready to automate the data side of the campaign. But before I could do this, I was requested to personalise the data targeting.
In Adobe Campaign, we work off a segmentation model which incorporates different value tier fields. Each field looks at different aspects of the customers betting behavior. (e.g we may want to use one value tier for gaming activity and another value tier for sports).
The values assigned to customers in these value tiers would be in numbers and in descending order from 1-10, e.g. 1 and 2 are the highest values and it gets lower and lower as we go down in numbers. Zero would seen as no value.
Historically, we always targeted off gaming value tiers, but this request came in to segment customers based off different value tiers and assign them their highest tier between the two. So if a customer has a value of 2 in sports but a value of 5 in gaming, I should assign that person value 2. Once the customer is assigned a value, Movable Ink rules within the email will recognise this and assign the relevant email creative towards them.
Whilst this was easy to understand in principle, it was harder to apply in practice and meant that I need to configure new data statements again in the flow.
However, after a lot of testing, I would eventually create a new data field that would assign a customer their highest value based of different value tiers. This field would be called Value_Tier.
As shown from the table below, these are the data fields and formulas that I needed to write to make this possible.
This is a picture representation of the workflow in how the CVT0 and GVT0 fields had to be created first to create the Value_Tier data fieldTo check whether these formulas worked, I previewed the data in Adobe Campaign as-well as extracted a CSV file. From the screenshots below, I was successfully satisfied.
Email Reports:
Lastly, once I was happy with the improved data logic, I implemented this into our email report tool to provide better visibility to stakeholders on what customers were being targeted.
I used the same code that I created for the RDC and modified it for this campaign with a purple design. The JavaScript for the Counts and the new table can be seen below:
6.New Campaign Process:
After completing all the implementations and data testing, I took a step back to review the execution process and communicate it to stakeholders to make sure everyone was clear. The new process would be as follows:
The data flow would be scheduled to run in the morning of the intended date every week.
Once scheduled, the data would flow though to the approval which is assigned to Promo Ops instead of the CRM team.
Whilst this happens a crediting file would be extracted and its location will be shown in the SNAP of the email report.
Automated email reports will also be sent to the other relevant stakeholders of the campaign to make sure transparency is maintained.
Once, Promo Ops credit the customers in the crediting file, they can then approve the campaign and all stakeholders will be sent an email telling them that the campaign has been approved, as seen below.
The emails are then sent and the process repeats the next week.
7.Benefits of automating campaign:
Automating or semi automating this campaign accomplished time efficiency and provided a new template of executing daily campaigns.
Specifically, the benefits of the new process and automation of this campaign for each team could be seen as follows:
The Data team spent less time governing the data and ticking off checklists every week for a campaign that rarely changes.
The Marketing CRM team didn't need to create new email templates for the same promotion leaving them more time to prepare for other campaigns.
Promo Ops receive the crediting file quicker to before
The commercial team spend less time as an intermediary between different stakeholders to execute a campaign successfully
Furthermore, a key finding from this project was learning how to automate email reports. From this learning, I implemented it for other workflows used in daily gaming campaigns. This saved the data team on average 20 hours a week spent on executing campaigns and allowed us to use our time more productively on other tasks.
Since this launched, over 20 campaigns have been automated, saving hours each week for all the teams involved.
留言