Mubby🦉

2.8K posts

Mubby🦉 banner
Mubby🦉

Mubby🦉

@MubbyOlu

7 years of Web2 & Web3 BD & marketing. The marketing piece your project is missing. Founder @krydenstudio Founder @web3brandrep BDs @VictusGlobal_

Katılım Eylül 2020
2.1K Takip Edilen2.5K Takipçiler
Sabitlenmiş Tweet
Mubby🦉
Mubby🦉@MubbyOlu·
Please Read This and Let’s Discuss I’ve worked with both Web3 and Web2 projects in marketing and sales, and honestly, one thing keeps standing out to me there’s a huge gap between how both worlds handle marketing and sales. Yes, they’re different, but there should still be a meeting point somewhere. To understand this better, I had conversations with ten Web3 founders, people actually building real use-case projects, not meme coins or NFT hype. And what I discovered shocked me. There’s a big disconnect between their marketing and their sales. That gap is affecting their conversions. Most of these projects focus on hype, visibility, and shilling on X (Twitter). But that doesn’t always lead to real onboarding or user retention. In Web2, things are different. After building visibility, they make sure to step into the real world. They use different onboarding systems to convert all that awareness into actual users. That’s what sales should be. This was something Binance understood well in the early days of 2020 to 2021 when we worked with them on P2P onboarding. It’s also what sets many Web2 companies apart. For example, even with all its success, Uber still goes into the streets to onboard people. YouTube, closer to 2024, was still driving branded buses across cities just to onboard more creators. Crazy, right? But it works. Now let’s talk about Web3. Many projects think hiring KOLs and ambassadors is enough. But let’s be honest, most venture capitalists don’t care about your number of followers. They care about traction and value. Because traction proves that the problem you are solving actually exists within the community. I’ve seen projects like @cipherowl, @shield_xyz, @Glue_AI, @dakota_xyz, and @datahiveai raise over 3 million dollars each, yet they don’t even have up to 5,000 followers. That alone shows that followers and engagement don’t always convert. It doesn’t mean VCs will want to support you. In fact, when your project has poor onboarding and user retention, venture capitalists will likely want a higher percentage in your company. Your onboarding strength determines how strong you’ll stand at the negotiation table. That’s one of the reasons I started @Web3Brandrep, a community that helps projects go beyond just marketing, to actually onboard and retain users who believe in what you’re building. If you’ve done your onboarding well and you’re looking for the right venture capital partners, check out @PanteraCapital, @FactionVC, @lightspeedvp, @recvcx, and @CryptomeriaCap. These ones value traction over noise. I’m also open to speaking with any project founder or individual who wants to understand this better. Let’s talk about it. Web3 can do better, and it starts with understanding this gap.
Mubby🦉 tweet media
English
19
12
130
4.2K
Mubby🦉
Mubby🦉@MubbyOlu·
Invisible AI
Français
0
0
0
31
Mubby🦉
Mubby🦉@MubbyOlu·
@OlgaMegorskaya @TolokaAI This is so nice. Please, @OlgaMegorskaya, my account got suspended yesterday due to some little mistake by me; however, the mistake was caused by the fact that there were two applications on Playstore which I downloaded. Please, how can I file an appeal to explain better?
Mubby🦉 tweet media
English
0
0
0
25
Olga Megorskaya
Olga Megorskaya@OlgaMegorskaya·
At @TolokaAI we have always believed that managing human efforts can (and should) be solved as an engineering task. Over the time, the approaches changed: - in the era of classical ML, managing large-scale, but simple and atomic crowdsourcing tasks relied on statistical methods and EM algorithms to weight human judgments in aggregation; - early GenAI era demanded much more focus on pre-screening of experts and peer review as a major method of quality control, which was hard to scale across projects; - modern Agentic capabilities allow to control every submitted task, no matter how complex it is, against pre-defined quality criteria. Our agentic LLM QA can use tools to fully investigate the submission: download webpages; process images, audio, video content; run python scripts, etc. What is more, it acts as an instant feedback loop, providing signal for retraining both QA agent and human experts. Read more details on how we benchmark the quality of LLM QA, what metrics do we care about the most, and what it has to do with Socrates - in our blog:
Toloka@TolokaAI

600,000 quality checks. One month. The bottleneck in annotation quality isn't your annotators. It's your QA system. Toloka's LLM QA runs on every submission. Deterministic Pass / Fail / Unable to verify, plus coaching feedback that loops directly back to the annotator. 87%+ accuracy on a benchmark biased toward hard cases. Full breakdown: toloka.ai/blog/llm-qa-sc…

English
1
0
1
24
Zino of Web3〽️⭕️
Zino of Web3〽️⭕️@onlyone_zino·
In a world where AI is everything, how do you stay secured? Join me tomorrow by 6PM UTC as we talk about how AI is being used in Cybersecurity. Do you think of it as the ULTIMATE THREAT or the PERFECT WEAPON? Set reminder with space link below 👇
Zino of Web3〽️⭕️ tweet media
English
17
1
33
251
Mubby🦉
Mubby🦉@MubbyOlu·
By the end of 2026, the only roles that will be totally free from AI threats will be: 1. Founders 2. People who know how to use AI (AI Generalists) Which of these categories do you fall into?
English
1
1
7
111
Mubby🦉
Mubby🦉@MubbyOlu·
@annakaz I really resonate with this; data businesses are security businesses. In what we’re building, people contribute data and earn tokens. It’s all about tying value to secure access, and this discussion is really inspiring me further.
English
0
0
1
21
Anna Kazlauskas
Anna Kazlauskas@annakaz·
If you run a data company then you also run a security company Your ability to charge for your data depends on your ability to gate access to that data
English
4
0
16
574
Olga Megorskaya
Olga Megorskaya@OlgaMegorskaya·
If you have a STEM PhD or research background, we are looking for you. Design research-level problems that train AI to reason through hard science. Verify solutions with Python and get paid for the depth of your expertise. $3k–$12k+ per 2-month project + referral bonuses for every qualified expert you bring in.
English
1
0
0
28
Manny ᵀᴹ
Manny ᵀᴹ@Liambams001·
@MubbyOlu I’m on it…done the necessary steps,just remaining to start recording Thank you for sharing once again
English
1
0
1
11
Mubby🦉
Mubby🦉@MubbyOlu·
Make money with Atlas 1/2 Debunking AI job myths! We talked about realistic ways to earn with AI our space yesterday. One opportunity is Atlas Capture. Sign up, take their training & test, and join their Discord community.
Mubby🦉 tweet media
Mubby🦉@MubbyOlu

You have been lied to about AI jobs. The said you can make $50/hr; is it true? Yes and No I have tested a lot of platforms; let's talk about how it works. Set a reminder for my upcoming space! x.com/i/spaces/1kKzD…

English
5
1
12
668