Improved transaction speed by 80% for customers placing very large value orders.

Conflict of priorities across stakeholders

S - Direct sales have 3 revenue streams - online portal, offline sales and ERP integration. ERP is toughest market to crack but has most retention rate. Last year, sales team won a large deal worth $500M for direct ERP integration service. Orders as high as $10M can be placed with thousand line items which would contain hundreds of thousands of SKUs. At the time, my product’s architecture supporting ERP integration could scale up to 100 items only which was 1/10th of customer requirement.

T - I was brought in to assess opportunity and define product changes to scale our APIs within 4 months.

Failure: Initial understanding and complexity of the problem gave a wrong estimate of timelines and did not consider other priorities based on country launch efforts of unified sales systems. Led to failing to deliver.

A -

- I talked to head of sales for that region to understand bigger picture. Deal was $500M revenue for Dell, and he said there's a larger market we need to tap into.

Research: I did my market research and interviews with key customers and global sales teams I learned that performance and accuracy of order notifications are customer’s top needs which meant that I had to scale our applications to accept large orders in one go.

Concluded that this capability would open up larger $8B market for Dell globally. Mainly end of quarter transactions are quite big.

1. Analyzed IDC and Gartner report on purchase trends

2. Sliced for online ERP users

3. Considered Dell’s fair share estimate in 3 years.

4. Research says that an average company’s overall information technology spending is typically 3–5% of their revenue.

Large enterprises in US

Market (Revenue) for large enterprises is $5T x 3% of revenue is typical spend on IT x 25% spend on computers x 30% Dell’s market share = $11B

$3B already with Dell, so incremental is $8B.

Brainstormed solution options with architects and engineers:

A) Branching logic for regular and large orders: Too complicated to maintain feature parity in future.

B) Manual hack to split large order into smaller chunks. Business or my manager was pushing us to place a manual hacky logic in place to split large orders into smaller manageable chunks that would have poor customer experience and incorrect reporting.

C) Single code base for regular and large orders.

Second, I collaborated with engineering architects to understand how we can separate our domain-specific resources such as item, shipment and pricing metadata into resource URLs to improve latency. I convinced them to focus on this problem as it opened up a broader market to us. They were currently working on building single click quote to order capability for small business customers.

Engg concerns - we may affect regular orders if we break metadata storage into resource URLs. We need to maintain parity performance for regular small size orders as well.

Product strategy - create scalable product that can accept small and large transaction in single API endpoint. Open up new market of $8B with large account orders (Untapped market for ERP exclusive customers x % PC customers x % Dell market share)

1. Break the quote and order monoliths so that quote object doesn’t increase its size linearly as line items increase.

2. Create separate item resources that have product and pricing metadata. Product, pricing with discounts, shipping details, lastly taxes and fees.

3. Improve network timeouts and buffer size limits wherever I see bottlenecks.

Third, I shared a detailed test plan to measure success to ensure we address the risks and keep stakeholders aligned. For simple orders, we wanted no drop in latency and performance. Business teams and customers were pleased with the transparency.

**Conflict:** stakeholder priorities were different: sales teams preferred quick solution to automate manual process without changing anything else. However, customer problem and long term opportunity was bigger, despite of only 5% of current revenue. My analysis showed that ERP customers are the toughest market for Dell to crack, secondly, ERP customers free up dollars to allow us to focus on consumer UX.

How did you win the $125M customer?

1. Shared current system limitations and reasons, with next steps.

2. Requested to continue using offline sales method till we bring large order capability.

R - Created lasting relationship with the initial customers who gave feedback. We reduced 80% latency for large orders to increase our TAM by $8B, won 2 more accounts afterwards, capturing $200M incremental revenue in initial 3 months itself

L - Important to focus on clarifying customer problem instead of sales push to quick delivery.