Data 360: How We Cut Lead Assignment From 20 Minutes to 20 Seconds
Data 360 gave our lead scoring models the full customer picture and changed how we route about 50,000 responses a day. Here's what broke, what we changed, and what you can borrow.
Every year, Salesforce redraws account territories and refreshes its sales routing rules, often just two weeks before they take effect. Under our old process, leads could follow outdated rules for up to six weeks afterward, and sellers fixed misrouted leads by hand. This year, every routing change went live the same day the new policies did.
Getting there meant redesigning our lead routing platform. Assignment happened across three systems, each region had heavily customized rules, and almost every inbound lead went to a sales rep. Live chats and phone calls got fast attention, but everything else landed in one big pile. Someone asking for a demo looked the same as someone who downloaded an article, so reps couldn’t tell serious buyers from casual browsers, and the best leads got lost in the crowd. We were using incomplete data to answer a critical question: is this person ready for a sales conversation?
Key Takeaways
How Data 360 powers scoring and routing
We started with the customer profile. Data 360 brings website activity, event attendance, email engagement, support history, and product usage into one profile, covering more than 150 million individuals at Salesforce. We covered that foundation in 3 Things We Learned About Using Data 360 To Unify Our Data and went deeper on the design in Architectural Lessons.
This time we put scoring and routing on top of that profile. Our in-house data science team built two models that drive routing. A lead score ranks each lead from 1 to 5 to show when it’s ready for sales outreach. A product interest score shows which of our products the person is most likely to buy, so the lead reaches a team with the right specialty.
How a response gets routed
Every response follows the same path in seconds. Data 360 matches it to a unified profile and, if needed, uses Hybrid Search to find the person’s company. An agent classifies their job level, the models score them, and configurable tables pick the right sales team. A final check filters out people who wouldn’t influence a purchase.
Each response ends up on one of three routes:
| Route | Who it’s for | Who acts |
|---|---|---|
| Sales team | High-scoring leads with clear product interest | A seller in Sales Cloud |
| Marketing nurture | Leads with interest but low readiness | Marketing campaigns |
| Agent-first outreach | Leads that meet the configured eligibility rules for geography and product | An agent begins outreach and hands the lead to a seller when it’s ready for sales |
Agent-first rules are set by geography, company size and product, so in-flight deals and key accounts stay with human sellers. The agent route means more follow-up on leads that would have sat untouched, and more capacity without adding headcount.
Routing rules the business can change
Thresholds live in configurable tables, not code. When our AMER SDR manager wanted a higher bar only for Tableau leads, they edited the table themselves, and the next Tableau lead was judged against the new threshold right away.
Faster assignment, simpler routing
- Lead assignment: 20 minutes to 20 seconds, a 97% reduction.
- Custom code: down 90%, from about 2 million lines.
- Routing changes: four weeks to days, an 85% reduction.
We saw these results within the first year of the new process running in production. The results come from the whole implementation, including our custom models and routing logic. Data 360 supplies the profile those pieces depend on, but it isn’t the only cause.
The same profile powers more than routing. One always-on email journey for Agentforce Foundations customers, built on product usage signals in Data 360, gets a 233% higher click-through rate than our average Marketing Cloud journey.
We’re not done. We keep tuning thresholds as campaigns and products change, so routing never falls behind the business.
What you can borrow
You don’t need our scale to use the pattern.
- Unify before you score. A score built on one system’s view inherits its blind spots.
- Define what “sales-ready” means for each team. Use lead scores and customer context to set assignment thresholds, then adjust them as business needs change.
- Give every lead a destination. Define the nurture and agent-first routes before you tune the sales route.
- Make rule changes cheap. If a routing change needs a code release, the logic drifts away from the business.
See how Data 360 connects your customer data
