December 2025. 3:47 p.m. on a Tuesday. I was halfway through a code review when our VP of demand generation posted in #gtm-ops:
Can we add professional email addresses to 800 LinkedIn Sales Navigator accounts before tomorrow at 9 am? Friday launch depends on it.
My first reaction wasn't panic. It was more like: okay, this is about to become an API integration problem.
The old fallback would have been ugly. Assign each SDR 80 accounts, tell them to open LinkedIn Sales Navigator, click through a handful of profiles, and maybe find email addresses using a professional email finder extension. Even if everyone dropped everything else, we would not get clean data by morning. Not at that volume. And honestly, even if we did, we would have no verification trail. That is how bad lists are born.
Here is the context: LinkedIn Sales Navigator is still our source of truth for who is in market. It is a map of accounts and buyers. But Sales Navigator is not an email service. It gives us names, roles, seniority, intent hints, and a LinkedIn profile URL. It does not hand us a clean company domain for every row, and it definitely does not resolve a verified professional email. That is why we had okki-go in our stack.
I have mixed feelings about AI-led prospecting products. Honestly? Many tools promise they will run your entire outbound motion, and they usually create more chaos than they remove. Okki-Go was different in one important way: it had an API-first design. We had bought it a month earlier, but the SDRs were using it as a lightweight browser lookup tool. Nobody on my side had wired it into our internal systems yet. The okki go api integration was still on my roadmap, waiting for a calm day. This was not a calm day.
So I opened the okki-go documentation with the same skepticism I bring to every vendor doc. What I found surprised me. The endpoint did not need a perfectly cleansed CRM or a decade of historical data. It needed a small list of identifiers, and it would do the rest through waterfall enrichment: check one source, then another, then another, and return a verdict.
I wrote a small script in an hour. Pull records from our temporary sheet, normalize the fields, batch them to the API, then write the results back. The first run failed. Of course it did. I sent a field named website_url where the API expected a company domain. A batch of 400 errors, a quick fix, and then the queue started moving.
What data is required to find email? Honestly, less than I expected
The question that followed us the rest of the week was simple: what data is required to find email? The short answer is not a huge custom object with 47 fields. For a typical B2B contact, we needed just a few anchors:
- A company domain. A domain resolves the organization. If you only have a LinkedIn company URL or a website field, normalize it first.
- A full name. First and last name are the minimum. If you have a middle name or maiden name or nickname, it helps disambiguate, but it is not required.
- A LinkedIn profile URL where possible. This one felt like cheating. When we sent a profile URL, match quality went up noticeably because okki-go could confirm it had the right human.
- Optional filters like title, country, or seniority. These were not required for email finding, but they eliminated wrong-person matches.
That list is the actual answer. If you are looking for what data is required to find email, start with company domain and first and last name. Everything after that is precision, not prerequisites.
People think a professional email finder magically scans the internet. That is a legacy assumption from the early days of email scraping. The modern reality is identity resolution plus verification. Okki-Go did not dazzle us with magic; it used the identifiers we gave it to find probable matches, then checked the syntax and mailbox status. The output was a verified email or a do not use flag. That nuance matters more than people realize.
By 8:40 p.m. we had gone through all 800 rows. If I remember correctly, okki-go returned a candidate email for 671 records. Its verification pass flagged 78 of those as risky. We suppressed those, plus another 11 duplicates that were already in other sequences. That left 582 usable contacts. I want to say 582, but do not quote me on the exact number. The point is we went from a blocked campaign to a clean file before our internal cutoff.
We did not send the next morning as one huge blast. The SDR team split the list by segment, wrote three versions of a first line, and launched at 9 a.m. I watched response data trickle in throughout the day. Not a miracle. Enough to keep the Friday launch moving.
What I would tell another GTM engineer
If you are evaluating okki go for gtm engineers, set up the okki go api integration before the fire drill. The product works well as a self-serve browser tool, but its real value in our stack appeared when it became a programmatic endpoint. I did not need to wake up a designer or ask an SDR to copy-paste anything. I needed a URL, a token, and a clear list of fields.
Here is the lesson I keep relearning: a professional email finder is only as good as the identity data you give it. More tools do not fix missing domains or wrong names. Okki-Go saved our evening because LinkedIn Sales Navigator gave us the map, and our own system still had enough domain-level data to feed the enrichment waterfall. In other words, the old fundamentals have not changed. You still need a well-defined target audience and a clean source list. What has changed is how fast we can move from target list to verified contacts.
Was this a perfect process? No. We worked late, made one 400-error mistake, and lost several hours because I had not read the API docs earlier. But the difference between the 2020 version of me and this one is simple: I now treat list generation as a data problem, not a manual research exercise. Okki-Go was the trigger in this story, but the actual unlock was deciding that API integration belongs in the engineering backlog, not in the nice-to-have pile.
Bottom line: the next time someone asks for 800 professional emails by tomorrow, your answer can be yes. Not by asking SDRs to stay up all night. By building the okki go api integration once and letting it do the heavy lifting. At least, that is what worked for us.
