Implementing data extraction
Accelerating Pay history setup from 10 days to 30 minutes
Product
— QuickBooks Online Payroll: #1 payroll software for small business owners to manage their business and pay employees, with a customer base of 600.000 users
To comply with my non-disclosure agreement, I have omitted and obfuscated confidential information in this case study. All information in this case study is my own and does not necessarily reflect the views of Intuit.
CHALLENGE
40% QuickBooks customers migrate from a different payroll provider.
Migrators have to enter their payroll history for current year, so QuickBooks knows what taxes to pay on their behalf.
Adding pay history is tedious and painful manual process, with a lot of room for mistake — it takes 43+ fields to set up just one employee with pay history.
Average time to finish setup for migrators is 10 days, while customers expect setup to take about an hour.
Currently, 17% of our customers don’t complete pay history step during setup (each 1% translates into $480K in company revenue).
WHAT CUSTOMERS ARE FEELING
“There needs to be an easier way to enter payroll history. It took a LONG time”
— Rocklin Physical Therapy
“I would prefer an option where you all import it from ADP. At a minimum, there should be some way to import results with the needed data”
— CCCM Retail
“It is too hard to add pay history. Why can't you just calculated it for me?”
— FMNava Services
EMERGING TECHNOLOGIES
We were able to leverage Intuit Technology teams to automate manual experience.
Classification — figuring out which provider customer is migrating from and what type of pay history report customer is uploading;
Data extraction — extracting data from report and automatically filling the database;
Data completeness — figuring out what info is missing and asking customer to complete missing part.
MY ROLE AND TEAM
I was a Lead designer on this project, while also partnering with different teams across the Intuit ecosystem. I worked closely with Document Understanding Portal team on data extraction and ML, with Futures team on data completeness, and with Design Systems on creating a pattern for document upload and storage.
USER RESEARCH
As a team we made a couple of trips to our Reno office to care agents office.
Currently, agents support our Elite customers and help them collect previous pay history. It takes 8-9 calls on average to collect correct pay history reports.
It was exciting to shadow calls with customers and listen to the pain points.
I led a design sprint with the care agents and the team to figure out customer journey and brainstorm on the ideas on how we can improve experience with the new technologies, which led me to...
How might we...
…explain customers where to find their past pay history documents?
…be transparent about the status of upload process?
…validate data entry from customer?
…explain errors in easy and correct way?
…allow customer to correct imported info?
…build trust and let customers feel confident?
CUSTOMER PROBLEM STATEMENT
I am a restaurant owner.
I am trying to switch my payroll from ADP to QuickBooks,
but I don’t know if I can,
because it takes too long to enter data and I’m afraid of making mistakes,
which makes me feel overwhelmed and worried that I won’t pay my team on time.
HYPOTHESIS
Automatically import data from pay history documents, creating a streamlined experience that asks for what’s missing will increase setup conversion and reduce overall setup time.
VISION
Together with my visual design partner I created a vision for pay history extraction to set North Star to the team and help prioritize the project during 2021 Payroll strategy planning.
(3:47 min)
GETTING TO THE IDEAL STATE
In order to get to the ideal state, together with the engineering team I broke down experience into doable milestones, prioritizing what information would be the most important for our customers to have, and hence — what we want to extract first.
FIRST EXPERIMENT
While Technology teams were working on the data extraction, I designed and launched dry test — one extra screen during setup to prove if customers are willing to upload pay history reports.
I was tracking how many customers uploaded report VS skipped the step, if they uploaded correct files, and if they needed help article. Customers, who invested in finding and uploading report, tend to complete Payroll setup. Customers, who selected “Other” as previous provider, lacked clear instructions where to get the report, so the upload rate was lower.
Also, I was tracking which top payroll providers customers are coming from.
Since every payroll provider has their own report format that Technology team had to process, we kicked off with extraction from TOP-4 providers that would give us 68% user coverage.
RESULTS
After I presented vision, initial learnings, and roadmap to Payroll leadership, Pay history project is prioritized for the next financial year. We are aiming for the high goal —