Robby, a better way to work retail

Act I - Shifts

My founding teammate Josh and I were thrilled to get into the Australian accelerator Startmate. The program began in January 2015, and at the time our company was called Shifts.

From Roster Genius to a company

Shifts grew out of a side project Josh had built while working at Apple Retail. Employees received their schedules through an internal website that barely worked on an iPhone and could not export shifts to a calendar. Frustrated, Josh built a site that scraped the scheduling data and turned it into an iCal file. He called it Roster Genius.

He made it for himself, released it to his store, and then opened it to everyone. Within a year it had spread organically to every Apple store in the world.

Legend has it that Apple sent global communications asking staff to stop using the website.

I was hooked. Something Josh had made to solve his own problem had become useful to employees across an entire global retail network. There was clearly something there.

Eight thousand people before launch

After months of research, we decided to build a company around the idea. Roster Genius had been offline for about a year, so we relaunched it with one condition: to use it again, people had to preregister for Shifts. A simple landing page explained what we were making. In the two months before Startmate, 8,000 shift workers signed up.

Our Startmate pitch was an employee-driven scheduling app. Shifts would give workers control over their schedules and, by extension, more control over their working lives.

Much of the back end existed before the program, so we invited roughly 900 beta testers into a version we called 0.9. We deliberately left out schedule scraping to learn whether the app was useful on its own. The response was split, so we quickly built the integration that could acquire an employee's schedule.

Champion users

Our growth strategy was to find a champion inside each store and turn that person into a power user. They would grant us access to their schedule, we would build the integration with them, and then release it to everyone at that company. Apple Retail was the obvious first network. We had the relationships, and within a week the first version was ready.

The other Shifts

The following Monday, a TechCrunch article about a different app called Shifts circulated through the Startmate list. At first we thought the article was about us and had simply used the wrong screenshots. It had not. The other product worked differently, but the name, visual identity, and even the typeface looked remarkably similar to what we had announced.

There is still some mystery around the timing, but dwelling on it would not help. Within a week, we had a new name and a new brand.

Act II - Meet Robby

We became Robby, and things immediately went nuts. Hundreds of people signed up every day. Activation climbed through the roof: 95% of registered users added their schedule almost instantly. Once the first power user appeared inside a store, the rest of the team tended to follow. Coverage around the preregistration campaign also pushed the product overseas.

The first App Store call

One Saturday morning, about a week after launch, we received a call from the United States. Someone from Apple App Review told us we were violating guideline 4.5.1 and that our current build would be removed from the store.

The problem was our targeting of Apple Retail employees and the feature that scraped their schedules into the app. If we removed it, Apple said, we could return. We built a workaround that let users photograph a schedule and send it to us for conversion. At the same time, messages kept arriving from customers asking why the app had disappeared. We shipped a replacement build in just over a week.

Seven days in review

Normally an App Store submission waited four or five days before a human reviewed it, then spent somewhere between ten minutes and an hour in review. Ours stayed there for seven days. On the seventh day, Apple Legal emailed to say our developer account had been banned permanently. There had been no warning, only an email full of legal language and nobody we could call.

We reached into our extended Apple network for advice, but nobody wants to become the person responsible for solving App Store problems. Responses took weeks. Opening another developer account and submitting the same product seemed likely to end the same way.

Android, the web, and one founder

Fortunately, we had built Robby with Cordova. Within days we had a respectable Android build and a web version for people who could no longer download the iOS app. Android was difficult to maintain across devices, the web experience was poor, and convincing Apple employees to use a non-native mobile app was not exactly easy.

The following week, Josh told me he was leaving. After the cease-and-desist letter, he could not see a path back. Without an iOS app, there was no company in his mind. The news was devastating. I wanted to stop immediately.

We had about 5,000 users who relied on us to retrieve their work schedules, and Startmate demo day was only two months away. Josh stayed for two weeks to hand over development and ship a few final features. Then it was just me.

Countdown to demo day

Five weeks before demo day, we tried to seed Robby into non-Apple stores through Android, but there was too little time. One week before the event, I submitted an iOS build under a new account. It was accepted. Nearly everyone redownloaded the app, and at least investors had something they could see in the store.

Demo day went well. We were a US company hoping to raise in the United States, so we knew we needed a US lead investor. The day before flying out, Apple removed us again. I considered staying home, but if there was ever a chance to repair the relationship, San Francisco seemed like the place to do it.

The first week in San Francisco was part Startmate, part jet lag, and part adjusting to StartupHouse. I took investor meetings while explaining why the product was not currently in the App Store. It was a terrible icebreaker, but an icebreaker nonetheless. Eventually a friend of a friend at Apple gave me one piece of advice: rebrand again.

Act III - Hello Atum

Within a week, Robby became Atum. A new build went to the App Store, Apple accepted it, and our users downloaded yet another app.

A fundraising catch-22

Several notable seed firms were interested. They loved the origin story, the reason we had built the company, and the traction inside Apple Retail. They also saw enormous platform risk.

Come back when you have sorted all of this out with Apple.

It was a soft no and a perfect catch-22. The product was growing and retention was strong, but everything depended on one email from Apple.

A chance meeting

On my final day in San Francisco, a partner at a seed fund I had been introduced to finally found time to meet. I pitched, and within 30 minutes we were discussing terms. It felt too good and much too fast. When he asked when I was returning to San Francisco, the answer changed from 'I am not' to 'as soon as possible.'

I arrived home to another email from Apple: we would be removed for a third time. After everything that had happened, I was almost relieved. I wanted to settle it once and for all. The best person to solve an Apple problem was probably Tim Cook, so I thought, fuck it, and emailed him directly.

A golden response, then silence

My email explained the situation and argued that, if Apple understood how much its retail employees valued the product, perhaps there was a way forward. To my surprise, people high inside Apple replied and asked to arrange a call. More names appeared on the thread. There seemed to be real internal interest.

The day before the meeting, Apple canceled without explanation. It was another huge blow in an already exhausting list. I was at my wit's end and ready to give up. Then, by some miracle, the app reappeared in the store.

Two weeks to close the round

I went back to investors. I needed money quickly. My savings and three credit cards were nearly exhausted, but I calculated that I could fund three more months of runway, flights and accommodation in San Francisco, and one week in New York to close the round.

I had nine venture meetings across two weeks. Most were interested because our prospective lead was supposedly investing $800,000. With a lead in place, the rest was domino money. By the end of the trip, roughly $600,000 was waiting on that first commitment.

But the lead became increasingly erratic. One conversation sounded as if we had never met; another sounded as if the round was already done. They shifted toward priced equity, replies became scarce, and I started to feel them fading. On my second-to-last day, the whole team invited me on a boat trip. Surely that was where we would close the deal.

Long story short, the boat trip never happened.

When the lead stopped answering altogether, I had to tell the rest of the round that the deal was no longer closing. We later explored selling the company, but nobody wants to buy a startup while it is visibly running out of road.

Six months

Everything in this story happened in roughly six months: three names, repeated App Store removals, a banned developer account, Josh's departure, international user growth, investor interest, a direct email to Tim Cook, and a financing round that vanished at the finish line.

And, to add salt to the wound, Apple eventually released its own internal shifts app for the retail team.