Manish Sinha

I attended my first enterprise b2b conference last week. Our goals were customer discovery & research. Here's what did and didn't work well.

What worked well:

• Post on communities announcing you're coming. I posted using a hashtag on linkedin and on a subreddit => 14 inbound => 4 IRL meetings

• Take every single opportunity to talk to someone. Going to a workshop?

• Talk to the person to the left and right of you. Going to lunch? Get multiple portions and go to different tables. Waiting in line? Talk to the people in it - they're as bored as you are +> 20 IRL convos => 5 valuable connections

• Anytime I participated in Q&A at the end of a talk, I'd preface my question with "we're a startup doing XYZ, <genuine question related to some part of the preface>". Easy way to "batsignal' to others in the room => 0 connections

• Asked a friend who runs a non-profit to hook me up with an org email which I used to register for the conference => 50% discount

• Printed business cards w/linked in QR code on the back => 4 given out => 4 people added me on LI (try to get them to scan on the spot)

• I always asked "is there anybody you're looking to meet, who should I keep on my radar for you?" to make interactions as transactional as possible => 2 people asked => 0 people I was able to help

• If the conversation is going well, I'll push for a phone number to keep in touch => 2 times asked => 2 times given

What could've gone better:

• Should've created a group chat and added all the inbound to it. Easy way to get even more exposure and offer value (eg this happy hour is going on here, this happy hour there). Lost opportunity to build a rapport with the sponsors who were hosting the events and would've liked another channel to push their events on.

• Should've taken more photos with people, tagged them, and posted them in the form of "Day X recap".

• Didn't have time to print a tshirt. Attempted to "peacock" with an anti ERP ERP sweatshirt, which had lukewarm results.

• Didn't go to any of the HHs or after parties. Maximized for time at conference. Probably more valuable to cut out early, rest, and then go to one of the later events.

Last week, I sat in on a session where the Director of AI Strategy @ Oracle talked about their vision for AI. Some notes -

• They realize that the learning curve for NS is high. They envision agents acting as an intermediary between the user and NS, avoiding the need for the user to learn the nuances of the product, rely on tribal knowledge.

• They acknowledged the barrier with extending NS, specifically mentioning having to reach out to IT to get custom dev done. They want to use AI to solve this problem, though they didn't go into how.

• Reporting building will go away as an administrative function, which makes sense given how well LLMs handle text->sql.

• They're using external LLMs, and were cagey when asked which ones.

• They demo'd a chat bot that can help rearrange charts, retrieve answers to questions about your data. Pretty shallow stuff engineering wise, but I think useful.

• Curiously (and beneficial to our own endeavor) - no talk of applying this philosophy to onboarding/deployments; entire focus is on live instances.