A client had started his business like most of us, just sending and receiving emails. Then he grew to using a CRM system for support, a different CRM system for live chat, another shopping cart system, and finally an unused email autoresponder service.
We were able to take raw database dumps, API calls, and plain old email inboxes to build a clear contact database that grouped people by company and gave fields for:
- Multiple Emails Per Person
- Multiple Phones Per Person (cell/fax/office/other)
- Multiple Products and when they were interested in them to determine buying cycle
- Contact Dates to help with buying cycle
- Multiple Names per person, along with nicknames and what they actually want to be called (Sam vs Samuel, etc)
- How much and when they spent on what products
- Other members of their company
Once we had this, we emailed in small groups with offers specific to them and eliminated about 20% of the emails that were bad (but no phone numbers yet).
Then we followed the emails with calls to find out who was still in interested.
We started with 232,158 emails and about 120,000 phone numbers. This turned into 93,245 groups (companies) and about 110,000 individuals.
After determining the buying cycle for each customer (we’d like to say “big data,” but it was really just some math and guesstimates). Now every customer gets an automated personal email during what we think their buying cycle is and the customer service rep gets a task to call them the same day.
Overall, it was about 300 man-hours of work, but the reorder rate went from 20% to 35% after implementing and it runs on auto-pilot now and forever.