Why you (yes, you, the CEO) should be talking directly to customers

In 2013, four years after co-founding Olark, I was waiting for some friends to order at a Starbucks in Mumbai when a woman approached me and commented on my Virginia Tech t-shirt. I was jet-lagged, but energized — being in a new place always gives me energy — and eager to connect with people. So, even though I was just barely finding my footing after only 8 hours in a city that couldn’t have been more different from Palo Alto, I struck up a conversation.

Talk about a small world. The woman, it turned out, was Aditi Avasthi, founder and CEO of embibe — an early Olark customer, and now one of the most successful e-learning companies in India. Our conversation ultimately led to an impromptu tour of Mumbai, including a post-Starbucks jet-lagged presentation at her company’s headquarters. Aditi and I talked at length about the challenge of building trust with students as she disrupted test prep for national exams in India; seeing her office, and meeting the passionate individuals who made up her small team, gave me so much more context about her business than I could ever have hoped to pull from a spreadsheet or graph. I was so focused on in the conversation that I arrived back at the hotel with only an hour to spare before the wedding I had flown to India to attend — my wife had started to think I might have been kidnapped.

At the embibe offices in Mumbai

Businesses exist to serve their customers — but whether you’re the CEO, middle management, or an individual contributor, it’s all too easy to get caught up in the operations and forget about the very real people who are buying and using your product. To avoid that, I’ve tried to spend as much time as possible talking with Olark customers over live chat, on the phone, and in person. These conversations keep me grounded in the realities of our customers’ day-to-day, and they also fuel my creativity as I think about how Olark can contribute to our customers’ goals for the future.

Businesses exist to serve their customers

Over the last two years, though, I found myself focusing more on building Olark’s internal organization, and less on the customer-facing side of the business. Many of the challenges we were facing as a company were more internal than external, and in the wider context of life (including becoming a new dad — my wife and I made the transition from 0 to 2 kids in just over two years!) I had to be somewhat selective about where I put my energy.

This spring, though, I returned from parental leave with a mission. I wanted to step back into the role of customer champion and re-ignite our focus on customers. Specifically, I wanted to go beyond providing amazing customer service (which our Customer Service team definitely has covered) to build a sense of connection with and passion for the people who we serve with our product. It was a big vision, but it actually boiled down to a simple process — talk to our customers, listen to their stories, and share their perspectives both within and outside of our organization.

So, I started inviting customers to chat. I hopped on the phone with a director of global customer service who had slipped in a hotel bathroom and was working from home while she healed. I talked to a VP of marketing who’d stopped working weekends so he could spend more time with his 17-year-old son. I took a verbal tour of a family-owned envelope factory, helped one customer optimize their automation rules, and learned that another cares deeply about response times because their business is built on turning prototypes around quickly. And in a very fun meta-discussion, I listened while a longtime friend and potential customer told me about the process of talking with more than 300 of his customers before launching their most recent app.

In case you stop reading here, I want you to know that these deep, open-ended, and occasionally rambling conversations have been one of the best things I’ve done as Olark’s CEO. My calls varied in length from 15 to 90 minutes, but every single one was filled with insights and context that I can use to motivate our team, drive product decisions, and keep myself energized and excited for the future.

If you’re leading a business, you should be setting aside time to talk to customers and understand their challenges — and not just in a narrow user research or customer service context. And if you’re not sure where to start…well, hang in there through a few more paragraphs, and I’ll tell you what worked (and didn’t) for me.

Recruiting customers

Conversations, of course, are two-sided — so once I set my sights on talking to customers, I had to find customers who were open to talking to me.

I started with a list of goals. I knew that I wanted to:

1. Talk with customers across different industries (e-commerce, hospitality, education, legal services, etc.)
2. Talk with customers of a variety of sizes (sole proprietor to large company, one chat agent to several dozen or more)
3. Go deep to understand how live chat fit into the bigger picture of the customer’s business — this meant that I needed customers to commit to at least a 20-minute call.

Then, I created an initial list of customers to reach out to. That list included:

• Our 500 largest customers
• Our 500 smallest customers
• Accounts recommended to me by our customer service team (essentially, anyone who they thought would appreciate a phone call, for any reason)
• A selection of customers who responded to our new account welcome email with a personal note or support question

I used google sheets to track my outreach and progress against my goals for the project.

Armed with this (sizable) list, I tried a few approaches to reach out and recruit customers. The first was an old-school “do stuff that doesn’t scale” approach — I combed through customer websites, explored LinkedIn, hand-picked individual contacts, and wrote personalized emails explaining my goals and inviting them to set up a call. My first attempt completely failed to get a response:

One of my first emails, which elicited radio silence.

I iterated on the email and tried again. I knew I couldn’t keep up the super-personal approach indefinitely, but taking more time with the first few recruits helped me hone in on an email style that resonated and earned responses.

The third iteration of my email finally led to a few great calls—however, I had realized by this time that even when I spent hours scrutinizing LinkedIn and our customer database, I was still mostly guessing at who to contact. So I stopped presuming that I could parse out every customer’s org chart, and started simply emailing multiple people at each company and asking for an introduction to the right person. This process was much more efficient, and with less time spent finding the right contact I was able to focus my energies on listening closely to our customers, sharing feedback with the team, and finding interesting customers to reach out to. I also began experimenting with subject lines and template copy that could be reused, with minor tweaks, across customers.

Eventually, I landed on a generalized email that captured most of what I’d found myself writing in each personalized message. At the scale I was emailing customers (tens at a time, not thousands) it appeared that the subject line did make a big impact, so I quickly adapted the subject line for each email before sending it off. Once I got into a rhythm, I found that I could send about 50–60 of these semi-personalized emails in about an hour.

This is the email template that I’m using now; I tailor it slightly to each customer.

Note: If you want to email all of your customers at once, here’s a great example email template from Groove. I’m keeping my messaging a little more customized for now, but I could see myself trying an approach like Alex’s in the future!

I’m still using this approach, and have seen about a 20–30% response rate (on a per-business basis), which I’m extremely happy with — especially given that one member of our marketing team, who I’ll allow to remain anonymous, bet me I’d never see response rates in the double digits. I still think there’s room for improvement, though. I’d like to block out time in my calendar and allow customers to self-schedule calls, for example, although I do wonder if a list of appointment slots will feel less personal. I’m continuing to experiment, and by the end of the year, I want to have a system in place where I can reliably talk to 2–4 customers a week without a lot of coordination.

Asking the right questions

At Olark, we have a great team of designers and PMs who regularly run user studies and product-focused interviews. I wanted my customer conversations to have a different focus. I was interested in improving my customer intuition, building a library of customer stories to motivate our team, gaining a more complete view of the people using our product and the challenges they face in their roles, and building long-term relationships.

I also wanted to have genuine, free-flowing conversations. So I went into the first few calls with a list of questions, but didn’t try too hard to keep things “on track” — if my first question sent us off on an hour-long tangent, that was fine, so long as I was learning and connecting with the customer. I noted the questions that elicited interesting responses, and iterated from there. At this point, I have a loose script that looks something like the following, although I still ad-lib quite a bit:

• Intro: I start the conversation by thanking the customer for their time and explaining that I’m here to listen and learn about their role, the challenges they’re facing, and how Olark helps or hinders them in overcoming those challenges. I intentionally try to set a personal tone — as we move through the standard “How are you?” small talk, I mention things happening in my family (e.g., I’ve been single-parenting my daughter while my wife is on the east coast helping her mom recover from a bike accident). Setting a personal tone and opening up beyond the standard “good, and you?” call and response helps build trust and makes for much more genuine conversations.
• Tell me about your role at [company]. This question helps me get a sense of how live chat fits into the company’s strategy and management structure. Usually, it also elicits information about the interviewee’s day-to-day routine and the skills and resources (e.g., IT support) that they do or don’t have access to.
• What are your goals for the quarter/year (or longer-term)? How do you know you had a good quarter? What performance indicators do you track and report? I want to understand what success looks like for this person and their team, so I can gauge how much we’re actually helping them and where we could provide better tools or information.
• What are your biggest challenges right now? When I ask this question, I emphasize that I’m looking for insight into their biggest challenges overall, not just with respect to Olark (I assure them we’ll talk about Olark’s shortcomings later!). Are they juggling a lot of competing priorities? Wearing many different hats? Struggling to get access to the human, technical, or strategic resources they need?
• What are your biggest challenges with Olark? If the interviewee’s challenges with Olark are related to their biggest challenges overall, I know we have an immediate opportunity to help them. Sometimes this means relaying feedback to our product team and advocating for a new feature, but it can also mean teaching the customer a few tricks while I have them on the call. For example, I’ve been able to solve problems for some customers just by walking them through a creative use of our API.
• What do you like about Olark? I’ll admit this is mostly a feel-good question, but whenever I get an unusual or unexpected answer, I’m reminded of the tremendous diversity in our customer base. People use Olark in all kinds of different ways, and features that we consider “nice to have” can be critical to a customer’s workflows.
• (for newer customers) How did you hear about us? When possible, I love hearing about the journey a customer took to find Olark, and this is also useful information for our Marketing team.

The questions above are solid conversation-starters, but they’re just that — starters. If a customer’s answer feels incomplete, or piques your curiosity, follow up. I’ve learned much, much more from followup questions that I have from my basic script — why one customer prefers us to a competitor, why another switched live chat providers after an internal reorg, how vacations and parental leave affected another’s workload and ability to take advantage of our software. I’ve also followed up via email, and asked customers to share artifacts (such as flow charts and spreadsheets) that have helped me better contextualize their feedback. Artifacts are invaluable for showing our team how Olark fits into a customer’s larger workflow.

Sharing insights

One of my goals for this project was to share what I learned from each interview with our entire team, ideally in a way that would generate excitement around customer stories. My first attempt at this was to simply post a link to my interview notes in Slack. That completely flopped — no one clicked the link, of course, because everyone’s busy, and I hadn’t done much to make it sound interesting.

After watching a couple of lonely links fade into the Slack archive, I realized I needed a better hook. I started writing up a short summary of each interview — almost like a teaser for the detailed notes — which I shared in our main Slack channel. I also created a new channel, #customers, where I shared the full notes doc and invited questions and comments from the team. Finally, I tried to share summaries and notes on Fridays, which tend to be more social and less “heads down” for most Olarkers.

So far, the elevator pitch has worked. I’ve seen engagement in Slack from across all of our major departments (engineering, product, marketing, CS, and HR), and have heard customer insights reflected back in meetings and presentations; a group of Olarkers even put together a “customer scavenger hunt” project at our annual company retreat, with no prompting from me. In short, after holding about 30 customers calls and sharing around 20 summaries with the team, I’m seeing a big impact on the company — and on my own energy and engagement.

Posts like these had much higher engagement than just linking notes to a channel.

Olarkers joining in these discussions included everyone from frontline support to backend DevOps.

What’s next?

I definitely have some product recommendations to make based on these interviews. But for me, the most important lesson has been that deep customer conversations are indispensable. These open-ended discussions yield insights and inspiration that you just can’t get anywhere else. My plan for the next few months is to continue a steady cadence of customer calls, refine my process, and set up opportunities for others at Olark to talk with customers, too.

In conclusion — I’d like to challenge you to pick up the phone and call one of your customers. Get to know them, ask about their work, and brainstorm solutions to their problems. Even if you go in with no script, no plan, and no goals aside from listening and learning, I guarantee you won’t regret it. Share summaries of your calls with your team — it can be incredibly motivating for them to understand the stories of the people your business is helping every day. And if you can’t think of a customer you’d like to talk to offhand, or if you find yourself digging to find 5 customer examples for a presentation, just start reaching out — you’ll be surprised by the impact it will have on you and your organization.

PS — If you have a tip, question, or comment related to customer conversations (or if you’re an Olark customer who’d like to schedule a chat!), I’d love to hear from you. Find me on Twitter @jaminben, or reach out on LinkedIn.