Notifications
Project Submitted for Job: Product Manager emp0302 2 by testkrish 0602
×
Advanced screening Submitted for Job: Product Manager emp0302 2 by testkrish 0602
×
Job Application for Job: Product Manager emp0302 2 by testkrish 0602
×
Advanced screening Submitted for Job: Ats Product Manager US - United States by testkrish 2201
×
Job Application for Job: Ats Product Manager US - United States by testkrish 2201
×
Mark all as read

Email Campaign List Management

Welcome to our Product Newsletter, a biweekly email highlighting top discussions, and learning resources for product managers.

What We Will Cover In This Edition:-

Top Discussions: 

1)  How often do you think you should talk to customers?

2) What’s your experience working as a backend PM?

3) How to manage the size of the email campaign list?

Top Learning Resources:

1. The growing specialization of Product Management

2. The founder dating playbook

3. What it feels like when you’ve found product-market fit

__________________________

Top Discussions

Question 1) How often do you think you should talk to customers?

I’m an aspiring PM, and the general sentiment I get from PMs I talk to is that you should be talking to customers frequently (though the timing varies). Then when I press them, almost 100% of the time each person admits that they don’t talk to their customers as often as they should.

I’d love to hear everybody’s experience here. How often should you talk to them, and how often do you really do it?

(I assume it’s different in B2C vs. B2B as well, but if that’s wrong please disabuse me of that notion)

– Juan Allo

Discussion

A] There is a lot of say-do in product practices. Just as there is a lot of say-do in humans/users/customers. What we say is not what we do. There are some product organizations that NEVER talk to customers but espouse human-centred and discovery practices.

You should “often.” That’s probably at least every couple of weeks. This is more to have a constant pulse and drip of insights. There will be a ramp-up of the intensity of the types of research depending on where you’re at in the discovery process for a given product/feature.

You could consider Jared Spool’s “exposure hours.” Being tapped into your users doesn’t always have to be you literally talking to customers. It could be recordings from user research sessions, looking over verbatim emails, surveys, etc. That said PMs should not completely outsource user insights.

Teresa Torres’s new book on Continuous Discovery is perhaps an “optimal” we might aspire to. In talks, she’s given related to her book she talks about the evolution of a lot of companies shifting to at least doing project-based research vs continuous. And that’s OK on the journey.

My experience is more in line more project-based research. Ebbs and flows. I did 10 user tests last week. I probably won’t do anything this week. Hopefully a handful next week. My most intense research effort in a B2B context was 20 interviews over 2 weeks + a few dozen user tests and accompanying surveys.

– Michael Yoffe

B] I’m a bit more forgiving when it comes to talking to customers. Talking to customers should be serving a specific purpose. It’s easy for new PMs to talk to one or two customers and take that as insight to act on.

The goal for product managers is to be the voice of the market, not the voice of the customer.

So, as long as you’re using customer interviews as a data point and as a tool in your toolkit to develop that market view that’s okay. In an organization where you might have tens of thousands of customers, you’re looking for more macro trends and I tend to rely more on usage data, patterns, and trends. Product managers have a lot of listening posts across a variety of tools and solutions these days to be close to customers and to aggregate findings. The same cannot be said for organizations that have few customers where PMs can easily maintain relationships with all customers on the books.

That said, I still try to tag along in Sales calls when we’re launching something new, Account Management meetings for major customers, the occasional customer onboarding call with the customer success team, and as many user research interviews with the UX team as possible. I tend to average around 5 or so external meetings per week despite a hectic back-to-back schedule.

– Richard Soneva

C] At  least once a week. Sometimes 3-5 times a week.

I made an experiment and it worked beautifully well: suggest your team (pick one that you meet very regularly) and suggest that you all spend the time you’d held a meeting and instead talk to a customer asking questions about experience, usage, success, you name it.

You’ll see how valuable it will be for everyone involved. But keep in mind that ideally, you shouldn’t have too many people on the same call to talk to a customer.

– Nathan Endicott

D] It depends on the org, target segment, B2B vs B2C etc.

I work for FAANG and UXR members are embedded into product teams. The team collaborates on a variety of consumer research e.g. foundational, concept testing, usability testing, and general research (day in the life of). All of these are then synthesized, quantified and then shared with product teams with specific recommendations. The teams can view the recordings, notes etc. to get a better understanding of the user.

It doesn’t matter if you have full-fledged research or you’re doing ad-hoc research, the key things you want to

  1. Validate Needs => result is prioritized JBTD statements
  2. Gaps in the current product
  3. Measure sentiment

– Dhiraj Mehta

__________________________

Question 2)  What’s your experience working as a backend PM?

I’m  a non-technical PM who recently started a new job only to find out there were no front-end components.

I have worked with APIs before, but I’d work with a front end connecting to an API, so I could view/test/learn what information is being passed, how the data’s used downstream etc. But now my product is microservice service to microservice service. Without being able to read the code, I feel like a burden to my engineers, unable to field basic support questions.

I’m not sure how to excel in my role, without customers to interview or competitors to research how can I find new opportunities or add value to my product?

I do get requirements from other dev teams but they code changes and so I prioritize them in a project manager fashion, but I’m just not sure if that’s what’s expected for this kind of role. I’d love to hear your thoughts.

– Lawrence Martin

Discussion

A] The API is your user interface – you need to ensure you know the business goals/use cases of your users, how the API is used (roughly), and what functionality they need.

The business goals and use cases are no different than “typical” PM work – meet with users and their stakeholders to ask them. Develop an understanding of their business.

To understand usage patterns, you have a few options. You can develop enough technical knowledge to intuit what API calls need to be made to satisfy the use cases you learn about. You can ask the users to literally walk you through their code structure (at a high level – you shouldn’t need to read this yourself). The info you want is usually what calls they are making, how many times, to achieve their goals – and how information from one call pipes into the next. Your engineering team may also be able to help you here by building metrics and maybe even records of sessions. Depends on how exactly your system is built and what it logs, but I think that would be possible.

Lastly, approach new feature requests like you would any other product. What additional data do users need to accomplish their jobs? You might not be able to intuitively handle ease of use questions, but people aren’t going to reintegrate just because you made something easier to use anyway.

– Nathan Endicott

B] Options:

  1. ask a developer to teach you how to use Postman, basically a free tool to call APIs. You can type your inputs and see the result.
  2. learn the absolute minimum of being able to read code so you can call the APIs yourself. Really it shouldn’t be that complicated, at least for some base cases. All you need are basic templates that you modify yourself depending on your needs (change the parameters basically)
  3. ask your engineers to build you a basic frontend for the APIs.

Everything else you can still do as a PM. Interview users, ask for improvements, pain points, prioritization, bug reports, etc.

I have a dual profile PM/dev so let me know if I can help. Happy to jump on a call and see what can be done.

Cheers!!

– Donovan O’kang

C] @Donovan, Wait, why would you recommend a front end for a product that is backend (API) driven? Unless users want that functionality, this would be a waste of time. The goal is to ensure their customers can use the API. If the product is an API it means it’s developers who use it.

– Carlos Dubois

D] @Carlos, There are many ways to build a basic interface for an API. Automated tools exist that can do it in two clicks.

But even done manually, It’s just a few fields for inputs, a “Run” button, and a textbox to display the result. It’s a few hours of work for a developer.

On the other hand, such a small thing might multiply OP’s impact on the team. If you can’t run basic queries on your main product your value to the dev team goes down dramatically.

Edit: such a tool could even be released to the clients as a troubleshooting/testing tool.

– Donovan O’kang

__________________________

Question 3) How to manage the size of the email campaign list?

Hi all,

How do you like to think about email campaign list size and its impact on ending up in the Gmail “Promotions” tab? for example let’s say you have a list size of 10,000, and you’ve already tested times of day, days of the week, subject lines, etc. And 70% of your recipients are Gmail users.

Would you send to all 10,000 in one swoop or drop smaller packages, let’s say 20 packages of 500 recipients each? Really curious.

– Rohit Kumar

Discussion

A] With Mailchimp, I’ve tested both and realized there’s no real difference in the open rate so I stick to one fell swoop. I used Zoho campaigns for a bit and they limit the number you can send to at any time and deliverability is pretty bad; so my advice from experience is that you choose an email platform with great deliverability.

– Nathan Endicott

B] IMO the best way out is to send emails that users will engage with often. You can try sending plain text (that typically gets higher open rates in my experiments).

But nothing beats emails that get your users to reply. So put as many nudges as you can to get them to reply.

Hope this helps.

– Michael Yoffe

C] We’ve been running an email campaign with HubSpot in a similar environment. We’ve found that using other terminology other than ‘View online’ and ‘Unsubscribe’ can really make a difference.

– Karan Trivedi

D] Wow! Thank you so much. Grateful for your responses @Nathan@Karan & @Michael.

I think using MailChimp would be a wise choice as it seems to be very popular, obviously with almost all the features needed for running a good campaign. Thanks, @Nathanendicott for sharing your experience. Am glad that it does not make much difference in shooting out emails to a large number of recipients.

So I think I will stick to that too. And the tip that @Karan shared is useful as well. Will avoid those terms in my emails. Thank you once again to all.

__________________________

Top Learning Resources

The growing specialization of Product Management

Most people looking to understand the Product world, and even early career PMs, think that there’s a standard type of Product Manager. They believe all Product Manager roles are created equal and successful Product Managers are experts in the same areas.

The founder dating playbook

As the next generation of company builders takes shape and new founder mafias spin out of places like Stripe, it stands to reason that more co-founder stories won’t fit the traditional mould, and will instead look a lot more like Lin’s. And judging from threads on Reddit, countless questions on Quora and all the “first dates” filling up coffee shops in SOMA, the carefully designed hunt for a co-founder is already afoot.

What it feels like when you’ve found product-market fit

Of these three milestones though, the most likely to kill your business is the first: not building something people want. And that is what this post focuses on — how do you know when you’ve built something people want? Below you’ll find stories from twenty-five companies — many of which have never been shared before — revealing the moment they realized they had something special.

__________________________

If you enjoyed this newsletter, please consider sharing it with a friend by asking them to sign up here.

Who’s Prowess? We are optimist product managers, engineers, and educators working on creating a world where merit meets opportunity. On Prowess, aspiring and experienced product managers showcase skills, learn from the community, and connect with employers to advance their careers.

How can you grow with Prowess?

    1. Learn from curated learning resources and community
    2. Work on curated projects or join expert-guided group projects
    3. Receive personalized feedback from product leaders
    4. Build a portfolio to stand out as a product manager
    5. Access and apply to curated jobs
    6. Prepare for interviews with Q&A from top companies

Latest Posts