Salesforce – Part Two

I needed a keyboard. 

I stood at my reserved desk on the 7th floor of Salesforce’s tower in Vancouver around 9am. There were a pair of monitors mounted to the desk, a power strip, and a chair. Even a mouse. But no keyboard.

The office was mostly empty, with a few people taking calls in phone booths. The Salesforce mascots leered vacantly from omnipresent wall murals. “TechForce is on the 5th floor,” the receptionist told me from behind her face mask when I asked where I could get a keyboard for my desk. 

I took the elevator down, badged back in, and found the TechForce area. Except TechForce is only open on Thursdays, I learned from the sign on the desk. It was Tuesday. Nobody seemed to be around.

I arrived that morning with the intention of making the best of our new situation. We’d been encouraged to come back to the office one or two days a week. This was before the Omicron variant of COVID-19 exploded across the world. We were deep into the second year of the pandemic. I was pretty sick of working alone at home and missed my colleagues. Figured it would be good to meet some of the local Salesforce people.

Slack was now fully acquired by Salesforce after a 6-month period of bureaucratic limbo following the purchase announcement. We were in the process of integrating into our new parent company in the fall of 2021. Our beloved Vancouver office remained closed, but the Salesforce office was open. Except TechForce. “Or is it Techforce?” I wondered to myself as I made my way back to the elevators. “Is the Force meant to be capitalized?” There were a lot of forces at Salesforce to be understood.

I just needed a keyboard.

The cultural gaps between Slack and our new parent were evident everywhere you looked. At Slack our tight-knit culture informed every aspect of our work; how the product was designed, how we communicated with customers, how we ran the service and sold it, and how we coordinated internally.

If you needed help at Slack, you asked in a channel and a human would help. Things got done quickly and courteously. If you needed help with something at Salesforce, you filed a ticket and got an automated email response, and most likely a further automated followup with a link to an internal web page. We understood the practical and logistical differences between our 4,000 person company and our 80,000 person acquirer. Ours was a small town. Theirs was a sprawling city. It was nevertheless a shock for us.

We now had to do mandatory training in Trailhead, Salesforce’s internal learning platform. This involved frequently answering dozens of multiple choice questions. They were easily gamed but gave Salesforce the audit trail they needed to ensure all employees were up to date on internal policies. New job-required quizzes arrived regularly in our inboxes, like spam you couldn’t block. And we had inboxes again. Salesforce used email for internal communication. Coming from a company where our CEO had joked that anyone sending an email for internal communication would be fired, this was an unwelcome reality.

Everyone kept welcoming us to the Ohana. “What the fuck is Ohana?” we wondered. “It’s cultural appropriation is what it is,” others answered bitterly. Salesforce used the Hawaiian word for family as a welcoming umbrella for the company and its employees. They didn’t shy away from using other symbols of Hawaiian culture either: leis and traditional Pacific Islander greetings are common at Salesforce events.

Genuine or not, it rankled many of us Slack people. Slack had certainly developed its own share of corporate cognitive overhead in our 8 years of existence. But what we had was nothing compared to the overbearing weight of the Salesforce monolith.

But these irritating issues were the least of our problems as a newly swallowed subsidiary. 

Stewart was brought in under Salesforce’s Bret Taylor, who soon after became co-CEO alongside Marc Benioff. The Slack deal had been made with the working assumption that Bret would be steering as operational CEO, Stewart could focus on making the most of the overlapping capabilities of Slack and Salesforce, and Marc would continue on as the figurehead and founding CEO.

From the get-go Stewart found it difficult to connect with Salesforce’s founder. “I only got about one hour of total facetime with Marc before the deal was announced. My last meeting with him was cut short because McConaughey arrived,” he told me. Matthew McConaughey, that is. Salesforce reportedly pays him $10M annually to be a spokesperson.

Once the deal closed and we began the real work of corporate integration Stewart became deeply frustrated with how poorly run the company seemed to be. Operational efficiency was considered a critical component of Slack’s success. We routinely spent time and energy on maintaining and improving that efficiency across all layers. These efforts were anchored in the experience of Slack’s employees and customers.

Could engineers make meaningful contributions reliably? Did internal events run smoothly? Were product releases coherent and well-considered? Did the cameras in meeting rooms work? Was customer feedback understood and acted upon? Did desks have keyboards? Everyone took accountability for their piece of the puzzle. And everyone felt a deep responsibility to each other to eliminate papercuts, to own and fix chronic issues, and to generally strive to do the best work of our lives.

At Salesforce, Stewart was baffled by the breadth and depth of operational inefficiency. He witnessed executives green-lighting high-priced celebrity appearances while making cuts to things like inter-office employee travel. He saw VPs overhiring by thousands of people, then laying off thousands in subsequent quarters. An annual event for tenured and executive staff named Laulima (Hawaiian for “many hands”) cost millions to fly everyone to Hawaii and put them up in fancy hotels.

Stewart gave feedback on these issues and proposed solutions to Bret and Marc repeatedly, but it seemed to fall on deaf ears.

The practical matters of integration were also going poorly. After the initial restructuring that eliminated duplicate roles, Slack’s core responsibilities were largely left alone. Product, Design and Engineering continued on more or less as they had been before the acquisition. Customer Support stayed within Slack’s control for the time-being. But many of the processes and cultural norms around our teams began to inevitably shift toward Salesforce’s expectations.

This was felt most immediately in Sales. Salesforce makes targeted products for business units and sells to leaders of a single function, usually sales or marketing. Slack makes an inherently cross-functional product that needs actual buy-in from an entire executive team to be successful. Salesforce sells dedicated purpose-built databases for managing information; Slack sells an open canvas for any and every department in a company to collaborate more effectively. The resulting DNA of the two go-to-market teams couldn’t have been more different.

“Salesforce got put onto Slack [the product], but weren’t really using it, especially in Sales. They didn’t understand the product. So they sold the products they understood and were easier to sell vertically. Slack didn’t fit the pattern,” said Cal Henderson. Salesforce account executives weren’t using Slack in their day-to-day beyond direct messages, and had a hard time articulating the value to customers. Many thought adding Slack to a deal created confusion and elongated the sales cycle. And worse, Slack wasn’t a “dashboard metric” for Salesforce reps – there was little incentive to sell.

At Slack, customer engagements were focused on making customers successful, rarely pushing seats or over-promising if it risked a negative experience and slower growth. To be successful, Slack reps had to find internal customer champions and collaborate with them in order to teach decision makers that how their employees communicated could become a competitive advantage. The job of our salespeople was to tell the story of organizational transformation and then help with the logistics of putting Slack in place.

In contrast, Salesforce was built on crafty marketing (“no software” is a great slogan for a software company), and old school sales tactics: build an executive relationship, sell a vision, land-and-expand from CRM to add-on modules, and implement with paid services. The goal was to sell as many seats as possible across a sprawling software stack that locked the customer in. 

Dave Macnee, one of Slack’s leaders in Customer Success, recalls the difference showing up in the way our new colleagues used Slack: 

“Remember when Salesforce people would @mention someone in a thread with no context as a way to force them to see something, like forwarding email? Aside from triggering a notification for everyone in the thread, which was crazy annoying, it also showed how Salesforce pushed info around by email, with senior people defaulting as more important. It was wild. There’d be a wall of @mentions in a thread. Just clear that the other side didn’t get the mental model or the emphasis on courtesy.”

Our companies made strange bedfellows.

On the product side, we were focused on how to make Slack the best tool it could be for remote teams. We had always supported this use case. Indeed it’s how the company started, with all of us distributed across San Francisco, New York and Vancouver and then eventually around the world. Slack was perfectly suited for the working arrangements that COVID necessitated.

But we were all struggling under the new normal of calendars full of painful 30 minute Zoom calls and greatly reduced time for in-person discussion and collaboration. Surely we could do better. Two internal prototypes in this direction were catching on quickly. 

One focused on impromptu audio-only chat, available with a click from a DM or channel. When a conversation was busy, or when written back and forth wasn’t resolving a topic, it offered an immediate way to switch to voice. This allowed people to do what they used to do in the office: go talk to someone. The tap on the shoulder or hallway chat that can get something figured out in a few quick minutes. The lack of scheduling and video were a benefit. Compared to another dreaded 30 minutes of staring at people in boxes on a video call, these informal audio-first calls were a relief. We called them Huddles.

Another prototype started with lightweight short-form videos in channels. What if people could share their daily update with a video clip instead of a message? It would allow a bit of visibility for people stuck at home, and could inject a little fun into people’s day. Maybe it could be used for more informal sharing from execs about what was on their mind. Initial prototypes were fun and sort of goofy. There was something we liked about them, but they felt a bit disconnected from the rest of the product. 

We iterated on this social-media inspired “stories” format for a while before landing on something more straightforward and useful: video messages. These were short recordings with associated tools that allowed updates to be shared asynchronously rather than in a scheduled meeting. It was great for demos, walkthroughs, and short presentations. They had the added benefit of allowing people to watch on their own schedule with a transcript, pause button, and playback speed controls. We found you could cancel a 30m meeting in favour of a few three-minute videos and a thread for followup discussion. We called these video posts Clips.

Many internal teams were switching meetings to this async video format, then using Huddles for informal chats to fill in the blanks. We tightened up the product feature set and iterated until we had something we were excited to share with our customers. Huddles and Clips went out in mid-2021 as companies around the world were settling in for the long haul of the pandemic.

We were thrilled with the customer response. As usual with successful Slack features, these new capabilities had been eagerly adopted and refined through the torrent of internal “dogfooding” feedback. By the time they got to customers, they were mature and became essential additions to the product.

This was our first major feature release under Salesforce. Benioff called out Huddles as an amazing product in a company all-hands and gushed about the speed at which we had developed and released it. Simultaneously, it was clear internally that Benioff himself didn’t use Slack and didn’t really get it. This combination led to a strange sense of uncertainty among company leaders. Were we the new golden child of Salesforce’s empire? Or an unwanted orphan?

By the fall of 2022, Stewart decided to leave the company. After leaving Flickr/Yahoo to start Tiny Speck in 2009, building Glitch, pivoting to Slack, growing the company from a tiny failed startup into a hugely successful publicly traded company, and finally selling to Salesforce for billions of dollars, he was ready to move on. It had been an eventful 13 years by any measure.

He shared that the tipping point came when he was invited to Davos to represent Slack and Salesforce. Thinking he would be there to talk to customers, he instead found himself scheduled on a panel about innovation with will.i.am, another of Salesforce’s celebrity spokespeople. “It was so fucked up,” he said. He was disillusioned at the disconnect between his leadership style and priorities and the role he was being asked to play.

Once acquired, Stewart fought for Slack’s culture and our product’s value within Salesforce. And he worked to maintain our independence and local leadership where it mattered most. But he could only do so much. “The problem has been, there’s no incorporation of the Slack culture into the Salesforce culture,” he said during a company all hands that leaked to the media. “And unless there is some element of that, then it’s not integration in any sense. It’s just elimination.”

He planned to announce his departure on December 1st. Lidiane Jones, a Salesforce VP, would succeed him as CEO. Our CPO Tamar Yehoshua would depart at the same time and tenured Slack product leader Noah Weiss would take her place. This changing of the guard was carefully planned to ensure a smooth transition.

On November 30th, the day before these planned comms were to go out, Bret Taylor abruptly announced his own departure. Both Slack and Salesforce staff at the company were shocked. The co-CEO who had brokered the deal to buy us was leaving, and chose the day before Stewart’s announcement to share it. He said he planned to “return to his entrepreneurial roots” and start something new.

“I had no idea.” Stewart posted briefly and preemptively, before being bombarded with questions about Bret’s sudden exit. He had not been briefed on this change.

What happened? It seems most likely that Bret made a play to take the sole CEO role at the company. Denied by Benioff, he resigned. After making a $28B bet on Slack, he was walking away. Those of us left at the company would have to make the best of it, without our founder and CEO and without the Salesforce co-CEO who thought we were worth a massive investment to bring inside of Salesforce.

I stuck around until the spring. There were projects to ship, responsibilities to hand off, and colleagues to help get promoted. But I was ready to go after 11 years spent building something I deeply loved.

Before leaving the company, I made one last attempt to meet up with my old Slack colleagues at Salesforce’s office in Vancouver. “I’m bringing bagels!” I posted in our #yvr channel. This garnered a few half-hearted reactions 👍😋🥯. By this point most people had resigned themselves to being fully remote. Few were making trips to the office unless their schedule necessitated a visit. And some had moved away from the city to more affordable areas outside Vancouver.

I arrived back on the 7th floor and went to the kitchen. I didn’t recognize any of the few people getting their morning coffee. I set out the bagels and cream cheese on the counter and toasted myself one. Working from my laptop nearby, I waited for someone I knew to arrive.

Eventually a couple Salesforce sales guys came through. “These bagels free?” one of them asked.

“Yeah go for it,” I said. “They’re yours now.”

Thank you to Dave Macnee, Kim Graves, Stewart Butterfield, Cal Henderson, and Myles Grant for their notes on this post.

Thank you to our readers for their patience on this second part of the Salesforce post. We have one more post to come. Stay tuned.