Coterie

Services: Organizational Excellence Strategy + Retained & Contingent Search (Executive & Tech)

Coterie champions the belief that the foundation of every successful company are its people & organizational excellence. As a female-owned business, we partner with founders building companies whose values are committed to creating meaningful impact, growth, and accountability.

We help our clients save tens of thousands of dollars, or more. Let's build together.

Founder, Retained & Contingent Search (Executive & Tech)

Hampton

I get to interview epic founders in my free time. What more can you ask for?

Membership

Trust Machines

We’re Trust Machines. We've raised $150M to build the largest ecosystem of Bitcoin apps. Trust Machines is doubling down on Bitcoin and Stacks to bring consumer apps with real traction to the Bitcoin space. We’re a startup that builds applications in-house with Stacks to unlock new experiences for Bitcoiners.

▪️ Drove the full spectrum of recruiting and selection activities.

▪️ Created & implemented talent acquisition strategies for all engineering & non-engineering hiring 
     globally.

▪️ Designed and drove the adoption of consistent and structured recruiting, interviewing and hiring 
     practices that supported a strong candidate experience and optimal hiring decisions.

▪️ Partnered with leaders and hiring managers within the organization to track the ongoing hiring needs 
     of their business.

▪️ Defined and drove the team to meet and exceed hiring benchmarks relating to time to fill, cost per 
     hire, diversity hiring, and hiring manager and candidate satisfaction.

▪️ Developed and executed end to end proactive sourcing strategies in multiple disciplines/domains and 
     partnered with external resources as needed.

Head of Global Talent Acquisition

Hiro Systems

Took a breather from leadership to work with some of the greatest people on the planet. Only left because I was given the opportunity to build out Muneeb’s newest brain child — Trust Machines. Couldn’t say no! 🧡

Hiro Systems has THREE 2022 Built In Best Places to Work Awards:

1) Best Places To Work NYC - All Sizes (#7 out of 100!) 
2) Best Places To Work NYC - Small Companies (#3 out of 50!) 
3) Best Paying Companies NYC - (#10 of 50!) 

Hiro (formerly Blockstack) builds developer tools for Stacks, the network that enables apps and smart contracts for Bitcoin. 

Hiro empowers developers with the tools they need to innovate and build on Bitcoin, in ways previously not possible. 

Funding Raised: $100M+ from Union Square Ventures, Y Combinator, Lux Capital, Winklevoss Capital, Naval Ravikant, other VC's & ICO.

SEC Approved: The Stacks Token (STX) is the first SEC approved regulated crypto token offering in the industry.

Lead Talent Acquisition Partner

Chief

Fastest female owned business to reach unicorn status. 🦄 👑 

Chief is a private network focused on connecting and supporting women leaders. Chief was founded to drive women forward into positions of power — and keep them there. 

At the current rate of change, it will take over 200 years for women to reach parity in business. We’re not waiting eight generations.

Funding Raised: $140M from Primary Ventures, GGV, and others. 

▪️ Title: Director
▪️ 98% Offer Acceptance Rate
▪️ 280% Chief Growth

Head of Talent Acquisition

ClassPass

"Recognized as one of Forbes' Next Billion Dollar Startups, ClassPass leverages proprietary technology to dynamically merchandise and surface over a million fitness classes for a seamless booking experience." (ClassPass became a 🦄)
General Catalyst, Google Ventures, Thrive Capital

A ‘Techstars New York’ company

95% Offer Acceptance Rate

Bank of America

Global Technology and Operations Division  - Technology Recruiter

Emerging Tech, GIS, COO, Architecture, Quartz
C++, Python, Java, Scala 
Emerging Tech, AI, Machine Learning, Blockchain, Cloud

VP, Director, MD hires
Proud partner of the BOA Diversity Recruitment Team

Selected to be converted to Vice President

Talent Acquisition

Goldman Sachs

GS STRATS

Quant, Technology, and Data Science Recruiter 

The professionals in this group develop quantitative and technical solutions in partnership with teams throughout Goldman Sachs. Their backgrounds include science, mathematics or engineering as well as economics and finance. In addition to traditional quantitative finance activities, such as pricing and risk management, Strats are applying their skills to areas that include corporate finance, regulatory reform, business intelligence and automated trading.

Radley James

Executive Recruitment - NYC's Top Banks & Hedge Funds

Executive Recruiter

Michael Page

Got fired from my first agency gig. Turned out ok. 🤷‍♀️ 

Reminder: your failures do not define you.

Recruitment Consultant

Founder at Coterie | Hampton | Ex: Chief, ClassPass, Goldman Sachs

Prior to running Coterie, Olga was the Head of Global Talent Acquisition at Trust Machines. A fin-tech company that raised a $150M Series A. She has 10+ years of experience building teams for multiple multi-billion dollar companies like Goldman Sachs, Classpass, Chief, and others. In addition, Olga is trained as an Emotional Mastery Coach by Alyssa Nobriga’s "Institute of Coaching Mastery" (An ICF Accredited Institution).

The US military named a programming language after her. 

She wrote the first published computer program.

And yet for 100 years everybody ignored her contributions to computer science.

I just discovered Ada Lovelace. Her story is fascinating.

Some background:

Only legitimate daughter of poet Lord Byron. She never met him. He abandoned her within a month of her birth, then died 8 years later in Greece.

So how could he have such a big influence on Ada’s life?

It happened in a strange way.

Her mom was SO terrified she would inherit her father's poetic madness that she pushed her into mathematics.

Remember, this was the 1800s. Women weren't lining up for calculus classes.

But Ada loved it. And started showing an explorative mind at an early age:

▪️At 12, she was designing a flying machine with wings and a steam-powered engine.
▪️At 17 she was discussing computer features (in the 19th century, yes) with an English polymath.

Meeting him changed Ada’s path.

At the time, Charles Babbage was working on a “Difference Engine”. Not exactly the most exciting thing to talk about with people at society parties. 

Friends nodded politely when he explained his calculating machine but Ada asked questions. Good ones.

They started writing letters.

And in those letters, something shifted. 

Babbage moved beyond his Difference Engine to something bigger. The Analytical Engine. Today we'd call it a general purpose computer. It never went beyond a concept but the ideas allowed Ada to see what Babbage couldn't. 

She made it clear just a few years later. 

It’s now 1842 and an Italian engineer writes an article about Babbage’s engine. Ada translates it, then adds seven detailed annotations that end up THREE TIMES LONGER than the original article. 

Note G stands out among all.

It contains an algorithm for calculating Bernoulli numbers. Many consider it as the first published computer program. It’s revolutionary.

But there’s more.

While Babbage saw the Analytical Engine as a tool for numerical calculations, Ada saw it as a tool to work with music, language, art. 

"The engine might compose elaborate and scientific pieces of music of any degree of complexity," she wrote.

This was 1843. Light bulbs didn't even exist yet!

Victorian society wasn't buying what a woman was selling about thinking machines so her work sat ignored for a century.

But ~200 years ago she was already asking questions we still ask about AI today: are machines capable of creating or do they just follow our orders?

Her life was riddled with health issues and a gambling habit that led her to use mathematical schemes at horse racing (!)

She lived fully, but for a short time. 

Ada died at 36, same age as her father, buried next to the man she never met.

She didn’t become a poet, but she approached learning as “poetical science”. 

It helped her anticipate how we see machines today: not just calculators, but amplifiers of human thought.

Olga Etkina

Chief Revenue Officer | Ex: Chief, ClassPass, Goldman Sachs

I’ve spent years chasing someone else’s career dreams. 

I don’t regret it, but there are a few things I wish I’d known.

Some of you know me from a past in talent acquisition, a path I fell into because it seemed logical, validated.

Like many, I fell into it. I was good at it, but I was also miserable.

I couldn’t separate the tendency to chase responsibilities from the need to listen to my needs.

I was trapped by a sense of responsibility that didn’t allow me to step back and see what was possible out there. 

In truth, I had no idea what my needs even were. 

All I knew was, things weren’t getting better with time. 

The more I kept climbing the ladder, the stronger the pull to pursue the next move.

I was following the perfect LinkedIn career trajectory, each role more prestigious than the last, but every Sunday night I felt the same dread. 

Every time someone mentioned the next natural ambitions (Chief People Officer) I wanted to burst into tears.

How can this be a good sign?

I started plotting an escape, or at least a reality where fulfillment didn't require misery.

I moved from career coaching to executive search, still searching for something I couldn't name. I was guiding others through their career paths while being lost on my own.

The daily routine still felt suffocating, but something was different.

It was the energy, the feeling of excitement coming from those conversations with the executives. 

For the first time in years, I felt a pull instead of a push. A signal cutting through all the noise. I should have trusted it immediately, but I didn't. 

Someone recently asked me if I regret those career decisions. If I’d do things differently.

It’s a difficult question to answer when you know you ended up in the right spot.

It took me some time to understand what "right" means, but I am finally not considering what’s next or how to escape. 

I am proud of the work we’re doing at Rosedale and excited for what we’re building (will share more in the coming weeks). 

So how can I regret the challenges when they pushed me here?

But if I could go back, I would tell myself to stress less and focus on building relationships. To listen to what the discomfort was teaching me instead of resisting the journey.

Maybe there was another way, but I'm grateful how things turned out.

Perhaps someone else needs to hear this in their career journey too.

I recently had lunch with a CEO of a $38M-revenue company who wouldn’t stop talking about his AI upsell tool.

We met at a café in Midtown Manhattan. 

I hadn’t seen someone this excited in a long time.  

He knows I’m helping companies implement AI, so he jumped right to his new toy.

“We’re adding about 8% to monthly revenue. We love it!”

He said it was quick to implement, and the returns were clear.

I’m wary of “quick” AI solutions, but I’m not surprised by the results. We expect even better returns for our clients.

AI implementations can be tricky though.

And I’ve seen the strong ROI of using AI for internal solutions as well.

I was curious. 

Was he using AI to cut the busywork from his sales team? 

I said, “You could get a better return over time if you also applied AI to how your team operates. Are you doing anything there?”

He shook his head. 

“Maybe in the future. I don’t see the ROI for internal solutions right now.”

Okay.

I’ve heard this before. 

He continued,

“I think it’s more about them having to get better at managing their processes.”

He “thinks.” 

But did he really know his team’s day-to-day?

The truth came out quickly.

“We’re under pressure. I can’t afford to spend a week mapping our internal processes.”

Ha!

This is EXACTLY what we do at Rosedale.

We consult with leadership and employees to see where AI can add value.

For the present, but also for the future.

The AI implementation is the big add, but the discovery phase quickly proves to be value-adding itself.

The CEO didn’t want to help his team because he didn’t know how much they needed support.

That same team was enabling the growth the AI tool was driving. But without relief from the repetitive work, the tool’s gains were at risk.

The tech worked, but the rest of the system couldn’t keep up. 

I see this too often now.

Poor change management is a common reason for no ROI from AI implementations.

Leaders don’t explain their priorities or don’t take the time to let AI add transformational value to their organization. 

AI can help you increase revenue now, but it can also help you rebuild processes for long-term revenue growth.

You don’t have to pick, but you do have to make the team part of the change.

P.S. At Rosedale, we use AI to build revenue-boosting workflows AND automation that clear the busywork holding teams back. In parallel. No either/or required.

“Is this going to mean layoffs?”

I’ve had 100+ conversations with employees on AI. 99% of them light up. This one was different.

Part of my current role at Rosedale is to speak with teams, understand their pain points, and show how AI tools can eliminate repetitive work so they can focus on what matters. 

You might think most people get defensive when they hear about AI.

I see the opposite!

When teams see how tools solve their problems, they buy in. This is a key part of any AI implementation, and (thank god) it happens more often than you might expect.

For all the doom out there, I notice a strong general excitement for what's coming.

But on this call, the employee came in on the attack. Within minutes, he asked if this partnership would cut jobs. For half the call, the energy was tense, almost combative. I could hear he feared for his job.

I know that feeling is rough.

Once I explained the real purpose (to make his work easier, not replace him), the conversation softened. He warmed up. The same person who started cross-armed ended up asking questions with genuine curiosity.

It was a new experience, but I’m not surprised. Dealing with pushback is part of my job. 

Sure, we pride ourselves on building AI tools that quickly deliver value, but we also want them to do so in the long term.

It’s why we act as partners, not providers. 

We can’t control how leaders communicate or what they prioritize, but we can share what we’ve learned.

There’s technical talent behind every tool we build, but there’s empathy behind how we deliver it. 

Lose that, and the changes won’t stick.

Sam Altman chose her to introduce GPT-5. Jensen Huang blocked an hour for her show. But Cleo Abram was underqualified for almost every job she's ever had.

So how did that become an advantage?

It all started in 2015.

While working at a digital consulting firm spun out of Obama’s campaign, Cleo was drawn to the journalism Vox was pioneering.

She loved the idea of breaking down complex topics for an audience. Vox felt like HER place, but she lacked the skills to be considered.

How do you get close to the hottest thing in journalism with

a) no experience
b) no related college activities?

If you’re stubborn, you look for any slightly open door.
So, that’s what she did.

Vox had a development team pitching shows to streamers, and she had consulting experience. She offered to help on the business side.
Her pitch worked.

Cleo became “the lowest-level person in the development world.” It wasn’t glamorous, it wasn’t about “journalism,” but Vox was THE right place at THE right time.

It’s 2018, their ‘Explained’ show lands on Netflix. It’s a big win for Vox and a success for Cleo’s team. But it didn't feel like it.

She wasn’t going to be involved in the show. Her part was done.

Cleo’s path is full of moments where she’s almost there, but not really there. Some people are happy to build a career close enough to what they want. She wasn’t.

There was a clear competence gap to fill and little time to do it. But she had her evenings. After work, Cleo enrolled in night classes on video editing, animation, After Effects. 

Sure, it wasn’t enough to become a producer, but it could help her get closer to the production space.

It did.

First came some late night video editing work for Vox. Then, season two of Explained.

An executive producer sent out an email asking for episode ideas. Cleo wasn't on the production team, but she was on that email list now.

She pitched. One of her ideas made it to Netflix. They approved it.

The Explained team could have said "She's not a producer, thank you for your idea" but they didn't. They brought her on to write the episode.

Then another.

Then she was hosting Glad You Asked.

Then came an Emmy nomination.

Each time, underqualified. Each time, she built the skills on the side to prove she could do it.

By 2022, Cleo had built enough momentum to leave Vox and start her own YouTube channel.

Today, CEOs of the leading AI companies seek her out because she's become "a translator between the people building the future and the people who will live in it."

"Every single thing that's been important in my career... beforehand, I'm like, 'Can I actually do this? I've never done anything like it before,'" Cleo says.

Being underqualified might not be a great place to be, but it can be a great position to start from.

“AI is replacing entire Customer Success teams”

No, it isn’t. But here’s how you can use AI to increase retention:

1) Onboarding center 
AI pulls client data from CRM, project tools, contracts, and usage logs into one shared hub. It pushes updates to relevant teams, sets milestones, and tracks progress. Builds trust, keeps momentum, reduces drop-off.

2) Churn risk alerts 
Connect HubSpot usage data, meeting notes, and email. AI flags low-engagement accounts with churn-risk language, posts them to a Slack CSM channel. CSMs act before contract renewal.

3) Lifecycle email lift 
AI pulls high-value customer segments from the CRM, looks at past email campaigns that drove repeat purchases, and drafts new subject lines and body copy matched to each segment’s behavior. Marketing reviews, edits, and sends.

4) Upsell timing suggestion
AI monitors campaign performance in Google Analytics, ad spend in Meta Ads, and project progress in Asana. When a client exceeds KPIs, AI notifies the account manager in Slack with expansion opportunities.

5) Client Intelligence
AI pulls leads, spend, and conversion data from HubSpot, Google Ads, Meta Ads, and Google Analytics. It compares results to the client’s targets, then suggests next steps (think shifting budget or pausing weak segments).

P.S. - We're helping B2B companies fix their retention issues with AI. If you're seeing 20%+ churn, DM me. Happy to share what is working.

I've now seen 9 execs review AI solutions with great ROI and do nothing. 
Here’s what’s happening.

For the first time I see the hype stopping people from taking action instead of helping them jump on it. It’s understandable.

I’m not sure if you’ve seen any recent AI demos, but marketing is getting more impressive than the service AI companies are selling.

(if you commented “WORKFLOW” under a LinkedIn post to get the “BEST AI UGC AD GENERATOR” and ended up disappointed, you’re not the only one)

Many executives are pretending they have AI figured out, but they don't. They don't understand the technology, so how can they know who to trust?

I know this because I talk to them every day. They're afraid of getting locked into something that wastes money and time.

Sure, some have followed the hype. But if they could go back, they’d implement AI in a different way. 

They wish they'd spent more time mapping out what they needed AI for. They chose a provider, not a partner. They went for the quick solution that promised more instead of the scalable one that could elevate their service.

Here’s the truth:

- AI implementations will take longer than most vendors promise.

- If you do it yourself, you’ll have to iterate with imperfect tools (make sure to have the patience for it)

- Managing change isn’t easy, but most of your team will welcome it. They just need to feel they’ll be part of it.

Choosing the right partner is hard, but skipping anyone who doesn't tell you the truth is a good place to start.

I spent 3 months interviewing leaders at $20-100M companies about their revenue problems.

The same seven themes kept coming up:

1) Churn prevention is not a thing. Analysis is messy and mostly reactive

2) Losing a client in the first 30 days is very common. Poor onboarding is a massive and recurring issue across service businesses of any size.

3) It’s not unusual for companies to lack (functional) systems to prioritize high-value prospects. Sales teams often spend just as much time chasing the wrong leads as the right ones

4) Founders often underestimate how much faster scaling can happen if they fix non-revenue issues early. These same problems resurface later but are more expensive to solve

5) Support volume is high but low-value. Customer service is inconsistent, slow. Upsell opportunities are often missed because stuck in email chains.

6) Reporting is reactive, not real-time. Leaders act on numbers that are buried and disconnected. No clear visibility on client health, no predictive capabilities on upsell opportunities.

7) Those struggling with growth are seeing AI as a possible solution to get out of stagnation. They’re more interested in using AI to add value or increase services for their customers rather than to improve their ways of working.

Anything else you'd add if you've been noticing the same patterns?

7 Revenue Insights:

Amazon tried using AI to write product pages.

It was a disaster... until they figured out how to fix it.

Here's what happened.

Amazon manages hundreds of millions of products. Every day, millions of listings need updating. The data (images, titles, descriptions) must be correct and optimized for customers to find what they need in seconds.

It’s a lot of work, so they started using LLMs to handle the task. 

The first results were... creative.

80% of AI-generated pages were unreliable. The system (they call it Catalog AI) gave an electric pump 15 horsepower (it didn't have any). It described sofa frames when customers wanted to know about the cushions.

It was a mess. 

But a helpful one. 

The initial struggles pushed them to develop a strong approach.

Here’s how they did it.

1) Initial AI + Human Audit
LLMs generated pages, humans scored each page’s reliability. 
80% were unusable, but they knew what to fix.

2) Simple Rules
AI rejects content breaking basic rules (e.g. weight without units).
No unnecessary changes (like "contemporary" → "modern"). 

3) Statistical Guardrails That Reason
They profiled millions of products to know what's normal. 

When AI suggests a 15HP electric pump, the system asks to explain the reasoning and often the AI catches its own mistake.

4) AI Checking AI
Two LLMs in conversation: one generates, one reviews. They’re trained on different datasets so they catch different errors.

The reviewer’s questions force the generator to analyze and (sometimes) abandon bad suggestions.

5) Test EVERYTHING with real customers
Not some changes. Not most changes. Every. Single. Change.

They track the entire customer journey: clicks, hovers, purchases. Testing millions of hypotheses every year (humans can manage thousands).

6) Learning Process
Every successful change teaches the AI what works. Humans analyze patterns to accelerate learning. The system gets smarter automatically.

An example of an early win: for one skincare product, Catalog AI dropped a long feature list for three benefits: "smoother skin, hydration, fewer wrinkles." Sales jumped.

It took time to implement the full system, but the results are there. Catalog AI went from 80% failure to 8% of changes already boosting sales. And it's still learning.

Now… 

Not all companies have Amazon infrastructure, but all can use early AI failures to build stronger solutions.

Every hallucination, rejected output, and failed A/B test becomes training material for the next iteration. 

AI implementations are complex acts of analysis and progression.

Early results are great, but structured failures can be even more valuable.

70% of a CRO’s week doesn’t make money. 

I’ve noticed it after talking to leaders at $20-100M companies.

As they scale, tasks change but the time waste doesn't.

Inefficient processes end up filling the CRO’s calendar so much that they start affecting the company’s growth.

Some examples:

▪️chasing numbers for board decks and forecasts
▪️stepping in to “save” at-risk accounts after the damage
▪️manually nudging deals and leads through the pipeline

Yes, these look CRO-specific, but they cost the whole organization every day: 

▪️slower to double down on winning channels
▪️less investment in retention initiatives
▪️heavier dependence on net-new acquisitions to hit targets

The CRO becomes a patch for system problems. Hours go to emergencies and reporting instead of high-value relationship building.

Here’s how we’re fixing this for our clients:

▪️Automated client reporting on what happened, why, and what to do with it
▪️Retention tools to track client engagement and trigger nurture sequences
▪️Smart lead sourcing for data enrichment and prioritization

AI won’t turn useless processes into valuable ones, but when used inside tested processes, it frees the CRO’s calendar for revenue work.

The teams most excited about AI aren't the ones you'd expect.

I spend a lot of time inside teams during AI audits and rollouts.

When I started doing it, I thought sales and client success would be the cautious ones. People talk about AI SDRs all day, so I figured they would feel exposed.

I was wrong.

They CAN’T WAIT for AI. They deal with crazy volume every day, so the idea of a tool that takes some of that manual work off their plate feels like relief.

AI can automate 70% of their workflows but they still own the last mile with the client. Maybe they know their roles will evolve, and are excited to lean more into the human side of their job.

It makes sense, no?

Where I do see pushback is finance and analytics. 

Sometimes it’s open hostility, sometimes it’s just worry in their questions.

And it’s interesting, because, sure, AI is starting to help in their world too. It can draft reports, pull data, build decent visuals. But to be useful, the inputs need to be clean and the processes tight, even a low error rate in financial models or reporting can create huge problems.

AI might get there eventually, but it’s not “I will make you worry about your job” kind of there.

So, where does this resistance come from?

A few hypotheses:

1) Finance people know that one wrong number can kill trust. They value speed, yes, automation, of course, but not if they come with more mistakes. The price is just too high.

Analytics controls the story behind the numbers. They are focused on process, they care about protecting meaning and method, they are not limited to the output.

2) Both care deeply about accuracy and context, so when AI shows up promising to help by skipping steps, it feels wrong. They've been burned by messy data before.

3) Most of them also aren't using AI tools daily, so their reactions are based on early attempts or assumptions rather than experience with what AI can and can't actually do.

I'm curious if this matches what you're seeing.

Are finance and analytics the cautious ones at your company too? Or am I missing something about why they push back harder than sales teams?

Would love to hear what resistance patterns you've noticed (especially if they're different from mine)

Most founders I talk to think their problems are unique.

They're not.

I was recently on a call with the head of client services at a $30M creative agency. 

His account managers are struggling to keep up with client requests. The team feels overwhelmed by processes that they can no longer handle so retention is becoming an issue. 

I've had the same conversations with founders at a $15M marketing agency and a $40M software company.

The details changed, but the underlying problems were the same.

Here's what I keep seeing:

Account managers are relationship experts. They know every client's goals, history, pain points, but when one asks about campaign performance, it takes them three days to build a report.

Three days.

By then, the client has already started wondering whether anyone cares about their account. 

The pattern is everywhere.

Companies spend thousands on new leads while existing clients walk away because no one can answer questions fast enough.

The B2B businesses we're building AI tools for understand this.

When a client asks about ROI, the system pulls campaign data, analyzes trends, and shows what's working. When engagement drops, our tool flags the account and suggests next steps before the renewal call.

This is what AI should be doing. Not replacing the human judgment or the relationship skills, but making those people faster and more effective in their interactions with clients.

P.S. If you're looking for AI that gives your team instant answers instead of dashboards to check, we're building tools that analyze your data and suggest next steps.

Zapier is categorizing employees by AI fluency.

100+ companies will copy it. 95% will see negative ROI.

Have you already seen this table?

It’s what Zapier’s CEO Wade Foster shared recently to show how they measure AI fluency internally.

The table is detailed. I haven’t come across anything similar before.

(not from a company this size)

Different jobs need different AI skills.

Individuals get categorized after a series of tests.

It’s great to see this level of transparency. 

But here’s what’s missing.

We only see the expectations.

We don’t see how they’re helping people get fluent with AI.

Sure, expectations are part of ANY AI implementation, but they are not enough.

AI initiatives fail when companies don’t have:

▪️psychological safety for sharing failures
▪️resources (budget and time) to upskill
▪️guidance on what tools and outcomes should be prioritized
▪️clear leadership buy-in
▪️enablement for peer learning

Zapier embraced AI before ChatGPT made it mainstream.

They've been building AI features into their automation platform for years.

They're probably doing it right.

But here's what even smaller companies can do to avoid AI initiatives failing:

▪️allocate every employee a monthly AI tools budget, no approvals
▪️assign AI support leads in each team for hands-on help
▪️publish simple, role-specific guides for preferred AI tools
▪️require leaders to demo successful and failed AI tests
▪️meet monthly to answer questions and share learnings

Tables can track competency, but culture drives adoption.

Running a business is a lot easier when you cut your churn by 15%.
 
Here’s how we are making it happen for our clients at Rosedale.
 
Churn prevention is STILL not a thing. I’ve noticed it after 100+ conversations with $5-50M businesses.
 
Yes, you might have a dashboard connected to a few data points.
 
One of your analysts might spend an hour here and there making sure it’s still working.
 
But how often does it get updated?
 
And how many data points are missing?
 
From what I’ve seen, churn analysis is still VERY manual.
 
It’s a lot of checking Salesforce, copy-pasting into Excel, tracking platform engagement and conversation history.
 
There are no alerts. No early warnings.
 
The realization that the customer has checked out often comes too late.
 
So, we are helping our clients fix this.
 
We’re building retention tools that integrate with Gmail, Salesforce, and your other platforms.
 
They analyze meetings, email response rates, and clients’ behavior to calculate risk scores.
 
It’s a system that automatically flags early disengagement signals before the client tunes out.
 
It works!
 
Churn drops by 10-15% and the same system can identify upsell opportunities.
 
Churn might be inevitable, but less churn is very possible.
 
You just need to notice it before it arrives.
 
P.S. If this resonates, DM me. I’ll show you how to spot churn signals 60 days ahead.

“Why should we even hire you?”
 
Since I’ve started talking to clients at Rosedale, I had to come up with a great way to tackle this question.
 
It’s not difficult, I know. You’re probably thinking, “but Olga, isn’t Rosedale providing so much value that hiring you should be a no-brainer?”
 
My answer is: we are.
 
We’re building tools that deliver measurable, clear returns. Some are proprietary, but others are tailored for our clients. They will own them forever!
 
But here’s the thing: we are *still* an external partner.
 
And if AI is this force that is disrupting business models (it is), shouldn’t companies start building their own AI teams?
 
Some should. But not all.
 
Here’s what companies should do to make AI implementations work:
 
1) Audit internal processes
 
2) Identify highest-impact opportunities
 
3) Define how to measure AI effectiveness
 
4) Define by when they will measure AI effectiveness (underrated part of the process)
 
But also:
 
5) Test models and prompts
 
6) Deploy functioning tools
 
7) Maintain the tools as AI evolves

 
All of this while running a massive change management effort.
 
Because you don’t want your teams to reject the new tools, do you?
 
So, yes, if you have the resources to manage this internally, go for it.
 
But if you don’t, find a reliable external partner and let them help you.
 
DISCLAIMER: I work for an AI implementation company. If you don’t trust my advice, here’s what our clients say about choosing an external partner:
 
▪️ Hiring is a pain. Hiring locally is a bigger pain. Hiring local people with deep AI expertise is just a nightmare

▪️We are a fractional resource. They like using Rosedale as a scalable solution to accelerate their AI adoption

▪️They have limited budgets / time / knowledge to come up with internal best practices, so they seek guidance from experts

And if you still have some doubts, here's some data from MIT's research:

I've had 80+ conversations with employees at companies implementing AI. What they tell me privately vs what execs think is... revealing.

Since joining Rosedale, I've guided $20M+ companies through the early stages of AI implementations.

I speak to leadership, but also employees. 

In these past few months, I’ve noticed there's a whole layer of resistance that can explain why AI rollouts fail. 

So I went through my notes to understand which themes are more frequent. 

Here are the 8 that keep coming up:

1) "We think you’re bringing AI to cut jobs"

2) "AI is making our jobs harder, not easier"

3) "We need humans in the loop"

4) "We don’t believe you care about our pain"

5) "We can’t keep up with the pace of change"

6) "You’re cutting off our learning path"

7) "We don’t see the value beyond cost savings"

8) "We don’t know how decisions get made with AI"

Most execs ignore them, but poor change management is one of the main reasons behind failures in AI implementations.

It sounds impossible, but employees’ resistance could be avoided with some simple communication tweaks.

Here some I’ve seen work:

1) Be explicit that AI is for augmentation, not replacement

2) Use AI to remove friction, not just push complex work downstream

3) Keep humans in the last mile of customer and team interactions

4) Show how AI makes daily work easier, not just how it helps revenue

5) Set clear rules and guardrails so change feels steady

6) Encourage entry-level learning opportunities and experimentation

7) Frame wins around faster work and better ideas, not only savings

8) Cut/Simplify messy processes before adding AI on top

9) Make decision-making transparent so people see how AI is used

P.S. Founders: if you believe any of these themes could be present in your company, please reach out. Happy to share how we're helping our clients tackle them.

I was on a discovery call last week with the Head of Account Management at a marketing agency.

We were discussing how they bring new clients through their process.

"After sales closes the deal," she explained, "I send the client our onboarding questionnaire to get their creative materials and budget allocation. Then our analytics team reaches out to collect their tracking requirements and data access."

She paused.

"Now that I'm saying this out loud, I realize that sales also asks about their budget during the sales process. And our tech team also needs data access for setup."

A longer pause.

"We don't really have a way to see what each other has collected. Everyone just... does their part."

I could tell she was realizing something, "clients do mention having to repeat information. We all need different pieces, but there's definitely overlap."

You won’t believe how often I hear this.

Whether it’s a boutique agency or a $100M one, client onboarding is broken everywhere.

It’s good for us, I guess, since we build AI tools that offer a single onboarding platform that keeps everyone aligned.

But how many clients are these companies losing RIGHT AT THE START?

The pattern is always the same.

Customers have to repeat themselves, key info gets lost in siloed conversations, different teams “heard” different things. There’s no single source of truth.

From the inside, it might just feel like friction, but if you step back you can really see how this impacts retention.

"I mean, onboarding does seem to take longer than it should. And sometimes halfway through, the client will mention something that makes me think they expected something different than what we're providing."

Of course clients question the vendor choice when this happens.

As someone who helps companies implement AI, I won't oversell what it can do. But this is exactly the kind of pain AI can solve.

(below an Harvard Business Review global survey about B2B client onboarding highlighting *the same exact* issues )

Most AI conversations focus on job displacement. But I'm seeing something different in the field.

I’ve been talking to founders and execs for months now. 

They all have two things in common:

- Their no.1 priority is using AI to drive more revenue / improve retention
- They want to hire fewer people, not let go of current ones

Does this mean layoffs won’t occur?

Of course not.

They’re happening now and will continue to happen (although blaming AI for all recent layoffs is superficial).

The focus stays on growth and margins, but proper AI implementations create efficiency that frees up value for great teams.

We’re seeing this every day at Rosedale. 

Companies are using our tools to increase revenue AND to make their teams more productive.

Great people benefit from this. 

They’re spending less time chasing information. They focus on the value-adding parts of customer relationships. They can think more clearly because they can find and organize knowledge more easily and quickly.

Sure, this doesn’t come by design.

Every AI project needs to make team productivity a main goal. 

But it’s happening. 

We’re just not talking about it enough.

—

P.S. Logan Kilpatrick leads Google AI Studio and Gemini API teams. 

He’s surrounded by great people, yes. But aren’t we all?

Female engineers are less willing to use AI coding assistants. A recent study might explain why.

A leading tech company rolled out an AI coding assistant to 28,698 engineers. After a year, adoption was low. Only 41% had tried it. Among women it was just 31%.

So what was going on?

HBR researchers ran an experiment with 1,026 engineers from the same company to see how AI use shaped perceptions of engineer competence.

It was a simple study.

The participants had to review the same Python code snippet. Sometimes they were told it was written with AI assistance, sometimes not.

Here’s what they found.

When reviewers believed AI was used, they rated the engineer’s competence 9% lower. The code never changed, only the perception of the person behind it.

And the effect wasn’t equal.

For women, perceptions of competence fell more sharply: reviewers rated them 13% lower when told they had used AI, compared to a 6% drop for men.

Here’s what I found even more worrying.

Male non-adopters were the harshest critics. They judged female colleagues 26% less competent than men for the same AI use.

The damage from this bias was already evident.

Follow-up surveys with 919 engineers showed many avoided AI entirely to protect their reputation. Those most scared of being judged (women and older engineers) were the least likely to adopt.

This is the bigger issue.

The bias is affecting how individuals embrace AI, but it's also sabotaging entire AI implementations. Employees avoid tools out of fear, licenses get wasted and the productivity gains disappear.

The good news is companies can prevent this.

Here’s what we’re seeing work:

- Run surveys or focus groups to see where bias shows up most
- Ask leaders and peers to share how they use AI in their daily work
- Judge output on speed, accuracy, and impact

I’ve seen these steps make people feel safer using AI. Without the fear, the tools get used, adoption grows, and the returns become more clear.

Taplio

Olga Etkina's Linkedin Analytics

Who is engaging with Olga

's Best Posts (last 30 days)

Want to drive more opportunities from LinkedIn?

Taplio

.css-1jgreyo{background-image:linear-gradient(284deg, rgba(39,43,72,1) 0%, rgba(33,47,183,1) 100%);color:transparent;-webkit-background-clip:text;background-clip:text;}Olga Etkina's Linkedin Analytics

Who is engaging with Olga

's Best Posts .css-qqfgvy{font-size:var(--chakra-fontSizes-md);}(last 30 days)

Want to drive more opportunities from LinkedIn?

Olga Etkina's Linkedin Analytics

's Best Posts (last 30 days)