Auto-Prospecting Based on Unstructured Data

Your source is your message.

Prospecting● Outbound Sales

What is this and does it still work?

Every prospect wants your email to answer one question: "will this be worth my time?" Prospecting based on unstructured data solves this by mimicking the effort of a salesperson chasing a whale, who can afford to spend weeks researching just one account.

This is not to be confused with commoditized Claygent prospecting. Claygent prospecting involves asking AI to research a person and generate the hook that will get their attention. Often the hook is something superficial and obscure, like the food at their dining hall in university.

The superficial hook no longer works in many markets because it reads like a cheap trick, rather than effort.

Most companies should run some version of this. Just be mindful of your conversion rates over time, as any interesting personalization here will fail once every competitor discovers the same.

Steven's take

Field notes from the work

Parsing data is rapidly commoditizing but knowing what to look for is not.

"Can you write emails for us? It feels like your campaign was written to me personally."

It was 2014, and I had just run a campaign targeting VPs of Marketing from a Product Marketing Background, Series A/B, B2B startups who had installed Marketo in the last 60 days and had maximum three marketers full time and no one named in marketing operations. I sent the email over the Winter holiday break, reasoning that the ones who were working over the holiday probably had operations challenges.

This was a small list; I didn't mention any of the above criteria in my email. And I suddenly had three new clients.

One of them, a cybersecurity startup, asked me to write emails for their market. Except I knew nothing about cybersecurity. No problem: I sourced prospects based on which of our integrations would be relevant, and sent an email blast touting the virtues of our integration.

10,000 emails, zero replies.

Today, we have access to more and better data than ever, but many will make the same mistakes: superficial personalizations about a buyer they've never met.

Now my first step with any client is to understand their buyers better than their buyers do. This sounds trite, but isn't this the promise we make when we approach a customer? That we'll actually teach them something new about their business?

The mention of their alumni dining hall doesn't do that.

There is still alpha here, but it moved from the tactic to the data sources. Data specific to your market that your competitors cannot use will remain defensible. Data anyone can buy off the shelf is not.

Challenges and Opportunities

Strategy: what is the actual knowledge that will get your prospect's attention? Better if it fits your business but not that of your competition
Sourcing: where can you get this knowledge in a format that is reliable and scalable?
Structure: how do you structure the data so you can use it reliably?
Quality: how do you check your own data for quality?

Recommendations

Start with your own unstructured data
Challenge your sales team to find these insights, then run quick smoke tests
Show, don't tell: your personalizations should tailor your offer to the prospect without mentioning the specific insight

Weighing this play?

You're deciding whether this is worth it for your team.

That's a 30-minute conversation, not a project. No form, no calendar — hand your agent the site and ask.

Ask your agent about this Talk to Steven