tarun odedra (@odedra_tarun) profile picture

AI | GTM

London, England
Joined September 2018
Photo by tarun odedra (@odedra_tarun)Photo by tarun odedra (@odedra_tarun)Photo by tarun odedra (@odedra_tarun)Photo by tarun odedra (@odedra_tarun)Photo by tarun odedra (@odedra_tarun)Photo by tarun odedra (@odedra_tarun)Photo by tarun odedra (@odedra_tarun)Photo by tarun odedra (@odedra_tarun)Photo by tarun odedra (@odedra_tarun)
I think the key metric in growth is what you fail at before you get to the experiment that works for you best.
1
53
tarun odedra retweeted
We used to build lots of apps in parallel to see what sticks. They all sucked. LLMs make features easy to build. But everyone has the same LLMs. Where you focus is where you win. Context management matters for humans too.
1
2
1
21
2,089
Write the one-page summary your champion forwards to their boss. The person on the call rarely signs. Their boss reads two paragraphs and a price, so write those yourself and send them before the call ends.
33
Add one free-text box to the demo request form: what made you book this today? The answers name the trigger, often a new hire or a board deadline. Sort them once a month and build the next campaign around the most common one.
34
Late-stage tech lets marketing claim the deals it sourced and sales claim the deals it closed. The same deal gets counted twice in the board pack. Both teams hit plan, and the company misses it.
31
Send the quote within an hour of the call. The buyer is still holding the problem they described to you. By the next morning it sits in a queue with everything else they have to approve.
29
Private equity cuts the free plan in the first year to lift conversion. The free plan was where next year's buyers were learning the product. The conversion rate goes up for a year, and then the pipeline thins out.
23
Offer the annual plan the week a customer hits their first result. At checkout they are buying a hope. A month in, they know what the product does for them, and the discount reads as a reward for staying.
1
52
Late-stage tech launches a second product when growth slows on the first. The first product's customers are usually still waiting on the fixes they asked for. Selling them a second thing is harder than finishing the first one.
2
44
Set the free trial to the day customers get their first result, plus a little. A fourteen-day trial on a product that pays off on day two leaves twelve days to forget about it. Find the median day of first result in your own data. Set the trial just past it.
1
35
My best post of the fortnight was seen by eight people. One of them replied, which put it top of my own ranking. The next two also had a single reply each. On a small account one reader moves the table, so I read it for direction and keep counting.
24
I posted 43 times in the last fortnight. 25 of those posts got nothing back at all.
1
33
The 43 posts drew 1,044 impressions between them, about 24 a post. The repost count was zero.
1
3
Method: 43 posts from 12 to 25 October, X's analytics read through Postiz. A post counts as getting something back if it drew a reply, bookmark, repost or like.
3
tarun odedra retweeted
I want to send you $1k a month Send me prospects - I’ll pay you 10% rev-share We closed a $10K MRR deal last week - that’s $1k/month for dude who sourced it @deskonesocial - we run marketing & growth for companies
2
1
8
397
I split my last 43 posts by pillar. The ones telling founders to do one thing this week scored 64 per thousand impressions. My private equity opinions scored 19, and five of the nine got nothing back. The opinions are the ones I enjoy writing most.
17
Write the onboarding email from the founder, and make it ask one question. What made you sign up today? The replies tell you which page sold the product, in the customer's words. Read them on Fridays.
2
2
31
Put the cancel button where customers can find it. The ones who hunt for it leave angry and tell people. The ones who find it fast often come back. The last thing they remember is that you were easy to deal with.
17