Dell/ Customer preference change caused increased complexity and forced us to innovate with dynamic boot sequence.

Difficult, challenging problem, taking risks

S - As senior PM, I developed multiple generations of high-performance PCs. In one such generation, 3-months before new generation launch of my product, we faced chip shortage issues during Covid back in 2021. Due to chip shortages, sales cae knocking on the door saying that new customers were moving from AMD CPU systems to my product and were demanding feature parity. Particularly for system bootup and storage technology.

This was significant opportunity for me. $90M+ contracts. Not having faster boot slot would mean we will lose the contract.

But as this was close to launch, I was initially hesitant thinking the demand is sporadic and would be very short-term. I was still open to being proven wrong.

A -

- To assess further, I asked sales to help with quick customer interviews to understand nuances and technical details. I learned that customers are looking for longer term contracts with Dell provided their technical needs are met.
    
- Spun up a spike with engineering to evaluate tradeoffs, it came down to motherboard spec updates which would be costly if done post-launch. Alternatively, factory could go through a training to bring new steps and changes to their assembly operations + software patch to enable dynamic spec supporting both Intel and AMD customer needs.
    
- Asked for an urgent leadership review to align Engg, Factory, GTM and Sales on new specs so that we can confirm to customers about upcoming product launches and secure contracts.
    

R - won contracts worth $90M+ which was more than 2% market share bump and saved $200K in complexity costs by avoiding a post-launch motherboard update.

---

Older story

Due to chip shortages, new customers were moving from AMD CPU systems to my product and started expecting similar features for system bootup and storage.

This was significant opportunity for me. $90M+ contracts. Not having faster boot slot would mean we will lose the contract.

Enabling faster storage meant 2 options: Change motherboard design or spin up new motherboard after 6 months.

It was time sensitive to offer this feature since customer bids were coming in 2 months.

A -

- Measured customer impact of the performance boost + mentions of storage benchmarks in sales bids + historical trends of technology adoptions in similar cases

- Opportunity could be worth $200M+ for my products.

- New tech adoption for gen4 boot drive was faster than earlier generations as per customer feedback. So offering this brought 2% increased sales from existing customers as well. Calculated 2% based on increase in revenue at higher adoption rate.

- Evaluated tradeoffs of offering feature with dev and factory new product launch teams

- Option 1 was to build new motherboard and rollout after launch. Option 2 was to innovate with dynamic installation to protect existing customer feature (RAID).

- Option 1 was costlier and took longer, had risk of losing customer. Prioritized dynamic boot feature because customer impact was higher here.

- RAID was a competing feature that was incompatible with faster storage. So dynamic boot tech was necessary.

- Negotiated with ops and engineering management on importance of this feature and customer impact.

- Ops pushback - we have never done this before on production line. Resisting change.

- Convinced that the change would be software level and workers will follow the on-screen instructions same as before.

R - able to overcome conflict for a customer benefit and saved net $200K in complexity costs plus we won the $90M contract after launch.

Also the software tool at factory opened up similar use-cases for other 10 products, expanding TAM by $500M on 2 adjacent products.

Storage benchmarks for performance - PCMark 8 Storage, Sequential Read/Write speed from IOMeter, SpecWorkstation Storage,

RAID - redundant array of independent disks - creates data redundancy and benefits from fault tolerance to protect highly valuable data in case of hardware failures.