Erin Abrams

Stop telling your team "It's a marathon, not a sprint." That advice implies that building happens at a steady, constant pace. But anyone who has actually shipped a product knows that is a lie. Real development is Variable Pacing. It is a Sprint, followed by a 10K, followed by a Half Marathon. Phase 1: The Prototype (The Sprint) Mode: Speed. Hard-coded variables. Zero process. Result: Massive visual progress. 0 to 1 instantly. Phase 2: The Alpha (The 10K) Mode: Structure. Real databases. Error handling. Result: Progress looks slower because you are paying the "reality tax." Phase 3: Production (The Half Marathon) Mode: Stability. Role-Based Access Control (RBAC). Security audits. Compliance. Result: The "Boring" work that actually makes the product sellable. The Trap: Your job is not to maintain a constant pace. Your job is to recognize which race you are currently running, resource it appropriately, and reset expectations when you cross the finish line of one to start the next.  Read the full article here on my substack: https://lnkd.in/g2gqNGpv

94

10 Comments

Davon Salter joins Foundry as Associate Director, US Lead Ops. His new role focuses on U.S. lead operations at a time when marketing teams are increasingly relying on data, automation, CRM platforms, and AI to manage and qualify demand. The appointment highlights the growing importance of marketing operations in connecting lead generation, data management, technology, and sales alignment. A notable career move within the evolving MarTech and B2B marketing landscape. Read More: https://lnkd.in/dvNjVvqt #MarTech #MarketingOperations #B2BMarketing #LeadGeneration #MarketingTechnology #AdTech #CareerMove

15

Stop waiting for "Tuesday" to book your holiday flights! The Google Flights team just crunched 5 years of data to find the best booking windows and trending destinations for 2026. 1️⃣ The "Magic Day" myth is busted, again Yes, Wednesday technically beat out Tuesday as the absolute cheapest day to book this year. But the difference between the cheapest and most expensive booking days? Just 1.4%. 2️⃣ What moves the needle? If you want real savings, focus on when you fly and how you fly, not when you press "buy": → Fly midweek: Departing Monday and returning Tuesday/Wednesday saves ~14% over weekend travel (and up to 20% domestically) → Take a connection: Choosing a layover saves an average of 20% over a non-stop flight 3️⃣ The 2026 holiday booking windows → Thanksgiving: Prices typically bottom out 34 days out (aim for October!) → Christmas: The sweet spot moved 5 days earlier this year: lowest prices are around 56 days out (late October to mid-November) Bottom line: Be flexible, book early, and don't let a "magic day" myth stall your plans. 👇 Check out the full breakdown and trending destinations on my post on the Google Blog: https://lnkd.in/eZyw7Uix

143

14 Comments

Recently we launched a new Product Builder role on EPD at Owner. The lines between engineering, product and design are blurring. The future is multi-hyphenate, high agency, and customer obsessed. A few things we have discovered along the way. Home Departments still Matter We are not yet fully collapsing these departments. There we still need a DRI for the problem, the solution and the execution. We will hire Product Builders in all three departments. They can span into other functions, but they will always cover one area in full. Excellence still Required We are not interested in folks that are a 5/10 in all three skill sets. You have to be 10/10 in at least one, ideally multiple! This is Inevitable Our PMs were already coding, our engineers were already designing screens. We are just naming and enabling the shift already underway. Are there risks? Sure. We don't want bad code or bad designs tossed over the fence by people who dont feel responsible for the results. But speed and excellence have always been hard to achieve. Let's build!

54

5 Comments

When I go to hire Product Marketing Managers into my team: I want hoppers. I'd hire someone with 3 years PMM experience at two different companies over someone with 5-6 years PMM experience at one company. Beating a dead horse 5 or 6 times isn't teaching you very much. Years of experience is overrated. Obviously this is generalized. if you're doing 6 years at one place because you're increasing responsibility (and title) regularly, that's different... and it's GREAT (it shows learning and growth). The point is that 5,6, 8... 10+ years of direct role experience means nothing to me. It's the context of learning and growth that matters. Direct years of experience is a very limiting and misguided filter to start with as a hiring manager. So... I'll look for hoppers. It's a great indicator of people who are serious about continuous improvement.

100

44 Comments

Actual pro-tip for applying to Headway or any other company on Ashby - if you have someone who can refer you, WAIT until you get that, don't just drop your resume. 99% of the time, you should NOT do that. Speed matters for fresh job postings, and 36 hours can mean the difference between making the first resume batch that a recruiter looks at, and the pile they never get to. It's also a great hedge in case your connection doesn't respond, or respond in time. Headway, however, uses Ashby for our applicant tracking system - and it will actually REJECT a referral if the resume is already in the system. Like, straight up tell the user that you cannot add this referral, form submission error. That is... a really weird process/design choice to say the least... And I actually think it's an Ashby issue. No other ATS I've ever worked with has done this. So this advice might also apply for any other company using that system. So - if you're applying to Headway, and thinking about messaging your friendly neighborhood connection-who-works-at-Headway (cough yes by the way Headway is hiring) versus submitting right away: this is the one case when you should wait.

43

5 Comments

It’s not your talent…It’s your operating model Most product led companies don’t fail because of bad talent. They fail because they outgrow the system that once worked. In smaller orgs, you can operate on alignment by proximity. Slack threads. Founder decisions. Roadmaps in someone’s head. As your org grows, that stops working. • Roadmaps become political instead of strategic • Teams overcommit because no one models real capacity • Forecasts rely on optimism, not shared assumptions • Decision rights get blurry • Leadership lacks a clean, consistent view of performance Nothing is “broken.” But everything feels harder than it should. Scaling doesn’t create chaos. It exposes system and operational weaknesses that were already there. The companies that scale cleanly don’t add more process. They design an operating system: Clear planning cadence Defined decision rights Capacity realism Simple, shared metrics Structure doesn’t slow growth. It protects it. At what team size did things start feeling harder in your org?

22

Six years in chemistry before I touched a product roadmap. I still consult through firms like BCG and McKinsey and directly to startups navigating personal care formulation, raw material supply chains, and international regulatory frameworks. Probably seems disconnected to people who know me as a PM. It isn't. Starting from zero in a field...really zero, no shortcuts - builds something that domain expertise alone doesn't: the ability to identify which problems are actually the same problem wearing different clothes. Supply chain optimization for a personal care brand and retention architecture for a SaaS product are not similar problems. The frameworks for thinking about both of them are nearly identical. That's not a coincidence. It's natural, and it’s readily discernible when you learn to think from structure and first principles.

8

In my first few months as Product Manager of Blinq's AI Notetaker, our team’s shipped a lot. But this launch I’m particularly proud of. AI Notetaker for phone calls. Built for people who frequently have meetings in-person AND by phone — think real estate agents, sales teams, small business owners, freelancers, and more. Call anyone from your Blinq contacts or device's phone book, capture the entire conversation (with consent), and never miss an action item because you didn’t have a pen. Every conversation, saved to the contact. 3 big milestones, 2 platforms, 1 new vendor, lots of testing, problem-solving and hustle. One week before we released, during a general user interview, a user asked for exactly this feature. How good. And we’re seeing great adoption so far! 🚀 Big kudos and congrats to the team!! Check it out here: https://lnkd.in/gaqX2Npc

70

11 Comments

I’m super excited to be attending the Ai4 - Artificial Intelligence Conferences this week. Anyone else in my product, engineering and design world joining? I’m curious to learn more about how companies are pushing products and ecomm to new heights with AI. There’s a lot of hype, but what are you doing with real world applications that are showing true results?

25

pinch me! I got to speak at Harvard Business School last week to a room full of MBA students. what a full circle moment. we talked about how AI is reshaping the PM role in two pretty distinct ways. first, the obvious one – productivity. new tools are changing how we manage projects, run workflows, and handle the everyday stuff. but the bigger shift is what we’re actually being asked to do now. PMs are going back to their roots as operators and builders, in companies big and small. we’re prototyping with AI, running our own data analysis, doing customer and market research without waiting on other teams. the job is expanding, and fast. so grateful to have been in that room. if you’re a PM not paying attention to this right now, the time is genuinely now. thank you Harvard Business School for having me!

82

10 Comments

We shipped an AI-assisted tool that was, by any reasonable measure, better than what existed before. More complete, more consistent. Less work for users. And then… people didn’t actually use it. At first, we assumed this was resistance to change or lack of training. It wasn’t. When we finally listened closely, the feedback was more like: “This doesn’t sound like how I would explain this.” The tool wasn’t wrong, it was just bit too confident, too polished, it collapsed nuance in a way that made users feel exposed, and not helped. Once we rewrote the output to be more tentative, leaving room for judgment instead of pretending certainty, usage went up. Your users don’t evaluate AI tools purely on correctness. They evaluate whether the tool preserves their agency. If it sounds like it’s replacing their thinking instead of supporting it, accuracy doesn’t matter. That changed how I think about “helpfulness” in AI systems.

13

1 Comment