Latest / B2B SaaS Talks with Fexingo: Enterprise Software, Sales Cycles, and Procurement / Why Enterprise Buyers Want a Data Lineage Audit
Transcript
- Lucas: So I've got a stat that stopped me cold this week. Gartner surveyed enterprise procurement teams back in March, and 43 percent of them now say they require a full data lineage audit before they'll sign a new software contract. Luna: 43 percent? That feels high. Two years ago that number was basically zero. Lucas: Exactly. In 2022 it was 11 percent, according to the same survey. So it's quadrupled in four years. And it's not just regulated industries anymore—I'm seeing this in mid-market manufacturing, logistics, even professional services. Luna: And if this conversation is useful for what you're building or running — Lucas and I do this show ad-free, and listener support is what keeps it that way. If today's episode was worth a coffee to you, that's the link: buy me a coffee dot com slash fexingo. Lucas: Yeah, genuinely — we hear from listeners who say this deep-dive stuff helps them prep for actual deals. So if that's you, we appreciate it. Now back to the lineage audit. Luna: So what exactly are buyers asking for? Because when I hear 'data lineage audit', I think — okay, maybe they want a diagram? But it's got to be more than that. Lucas: It is. There are really three tiers that are starting to show up in RFPs. The first is infrastructure lineage — which cloud provider you use, where the data centers are physically located, and who has administrative access. The second is application lineage — how data moves between your modules or your microservices. Lucas: And the third is the one that trips most vendors up: data pipeline lineage. That's the full chain from ingestion — say, a customer uploads a CSV — through transformation, storage, archival, and deletion. And they want that mapped with timestamps and access logs. Luna: So SOC 2 Type II doesn't cover this? Because a lot of vendors I talk to lean on that as their compliance blanket. Lucas: It covers controls — do you have a policy, do you test it — but it doesn't trace actual data flow. I've had a vendor tell me, 'We're SOC 2 compliant, we're fine.' And the buyer's response was, 'Great, now show me exactly where our data goes when a user exports a report.' And they couldn't. Luna: I've seen this stall a deal. There's a mid-market SaaS company I know — they sell a workforce analytics tool — and they lost three months in a procurement cycle because their engineering team had to reconstruct the data flow after the fact. They'd never documented it. Lucas: And that's the trap. Most SaaS companies build fast and don't map data flows until a customer asks. By then, you're scrambling, and the buyer reads that as risk. If you can't show them a lineage diagram in week one, you're already behind. Luna: Is this just an enterprise tax on smaller vendors? Because a startup with five engineers doesn't have someone to maintain a data lineage map full-time. Lucas: It feels like one, but Gartner's data suggests it's becoming table stakes. If 43 percent of buyers require it, and you can't provide it, you're shutting yourself out of almost half the market. But there are lightweight ways to do it — automated tools that generate lineage from your cloud logs. You don't need a full-time data governance team. Lucas: The companies that treat this as a competitive differentiator are the ones building lineage into their product — like a feature that lets a customer self-serve a data map for their own tenant. That actually reduces support tickets and builds trust. Luna: I want to get tactical. If I'm a head of sales or a VP of Customer Success listening, and my company doesn't have a lineage audit ready, what's the fastest way to get there? Lucas: Start the day a deal enters stage two. Not at signing. Stage two. Assign one engineer to map the data flow for that prospect's use case. It doesn't have to be perfect — it can be a whiteboard diagram that you then validate with logs. The act of doing it forces you to understand your own architecture. Luna: And once you have it, do you share it proactively or only when asked? Lucas: Proactively, but carefully. Include it in your security packet — a one-pager that shows the high-level flow. If you wait to be asked, you're already defensive. If you lead with it, you signal, 'We've thought about this. We know where your data lives.' Luna: That makes sense. And I think the buyers appreciate it because it saves them time. They can check the box without having to schedule a follow-up call. Lucas: Exactly. And the savvier buyers actually use the lineage map to identify what they call 'data concentration risk' — if all their data flows through a single third-party sub-processor, that's a single point of failure. They might ask you to add redundancy or a fallback. Luna: So the audit isn't just a check — it's a negotiation lever. 'We see you use AWS in us-east-1 only. We need a failover region.' That becomes a contract term. Lucas: Right. And if you haven't mapped it, you don't even know what you're agreeing to. I've seen vendors sign data processing agreements that require them to maintain data residency in three regions, but their infrastructure only supports two. That's a breach waiting to happen. Luna: Okay, so what about the data itself? Are buyers asking for lineage of derived data — like, after their data is run through your ML model, where do the model weights go? Lucas: That's the cutting edge. Most lineage audits today stop at the application layer. But in the last six months, I've seen two RFPs that specifically asked about model lineage — 'If you train a model on our data, is that model later used for other customers?' That's a huge question for ai first SaaS companies. Lucas: And the honest answer for a lot of vendors is, 'We don't know.' Because data scientists often train models on shared infrastructure, and the lineage isn't tracked. That's going to become a major blocker as more AI features land in enterprise software. Luna: So the takeaway for me is: if you're selling B2B SaaS, and especially if you're mid-market or enterprise, start mapping your data flows now. Don't wait for a customer to ask. Lucas: Absolutely. And think about it as a product feature, not a compliance burden. The vendors that make lineage self-serve for their customers will win trust faster. The ones that scramble to produce a PDF after the deal is already in legal review? They'll lose deals they could have won. Luna: Alright, I'm convinced. Next episode, I want to dig into how buyers are actually auditing those lineage claims — do they send their own tech team to verify? Lucas: That's a great angle. Some are starting to. I've got a story about a financial services firm that ran a penetration test against a vendor's data pipeline. They found a misconfigured S3 bucket within the first hour. That vendor didn't close the deal. Luna: Oof. Okay, that's a teaser for next time. Thanks, Lucas. Lucas: Thanks, Luna. See you next episode.