A little under a week ago I heard the news that Hubbub, the company I’d been working at for most of the last six years was going to be closing down. That time has brought with it the best job I’ve ever had, and that I wouldn’t trade for the world even if it didn’t work out in the end. This is story of those six years.
In May of 2011, after ten years experience as a software developer, I was looking to get out of freelancing as a web developer and into a permanent job after one too many ordeals trying to get paid for work I’d done (I never did get paid for that job). I wasn’t really sure where I wanted to work, but I did know that I wanted to work somewhere I’d actually care about what I was doing. I was 27 years old, and I was sick of jobs where my main role was to help companies sell people things they didn’t want, and didn’t need.
Frustratingly it's hard to distinguish between one startup job ad and another. Everyone is changing the world, doing revolutionary things, and they’re going to make a fortune doing it. I was signed up to all the job boards and getting a steady stream of emails saying very similar things. On the 24th I received one that included this line:
I am contacting you as you might be interested in running and growing a unique, award winning Ruby on Rails website for our small and fast growing start-up, Hubbub. Hubbub is the first home delivery service from outstanding independent food shops, and I am its founder and Managing Director. […] We offer a good salary, a fun working environment, and the chance to be part of something really big. Oh, and a staff discount on the best food in London.
Sounds good. The salary was in reality not that good considering the state of the market for Ruby developers at the time, but I love good food, so I fired off an application with the caveat that I’d want to work from home most of the time, and then promptly forgot about it. A fortnight later I got a reply, asking me if I could come in and do an interview.
That interview was the first time I met Marisa Leaf, human rights barrister turned tech startup founder. Even in that first meeting it was clear that Marisa was absolutely determined to make life better for independent food shops by giving them a chance to compete with the big chains who can afford to be open late into the evening. Also present was one of their early investors, Adam Cleevely, who was there to assess my technical skills. It went much how you’d expect a 2011 startup interview to go — a chat about my previous jobs, a few brain teasers, some questions on database design, and some questions to assess culture fit. I thought I’d done pretty well, but was a little wary about coming in as the company’s first in-house developer and taking over an application built by an agency. I’ve worked for agencies, I know what the software they deliver tends to be like!
Nonetheless, when I got an email the next day asking for references, followed a few days later by an offer to join the company I decided to go for it over a job at an agency. I’d be working for on something that seemed to be having a positive impact on local businesses, and I’d negotiated to work 4 days a week, 3 of them from home. Christmas would never be the same again.
I’m going to take this opportunity to describe how Hubbub works, apart from a few tweaks it remained fundamentally the same throughout. Customers would sign in to the website and select products from a range of local food shops; we had grocers, chocolatiers, butchers, bakers, and a bunch of others, although sadly we never had a candlestick maker. They’d then fill their basket, book a delivery slot, and check out. On the morning of delivery the shops would sign in to a web application to find out what had been ordered, pack those orders, and then report how much everything cost (there could be changes from the listed price due to weight variance, or substitutions). Around 2pm the delivery driver would set out to start picking up orders from each shop, and then deliver them in one go to the customer.
Written down like that it seems deceptively simple recipe. A big dollop of online shopping, two glugs of logistics, and a dash of customer service. Collect the ingredients, mix in the back of a van, leave to sit, and serve a few hours later. In practice there are approximately a million things that can go wrong along the way. Customers can’t find what they’re looking for on the website, shops run out of stock and send something wildly inappropriate instead, drivers break down during deliveries, a bug causes deliveries to be silently cancelled, people aren’t in when you come to deliver to them, a water main bursts and brings traffic to a standstill in Hackney. Those were all real problems that we had to deal with at one time or another.
The pay off is great though. Customers get really good food, and the shops get a much needed income boost from people who wouldn’t usually be walking through their doors. At least in theory a company that can make both sides of the deal happy should do well, and we came so close to doing so!
The Early Days — 2011
My first year was a blur, I’d done some freelance work prototyping startups before but I’d never properly been inside one. I was hooked.
Back in those days there were four of us in the office and two drivers. Marisa and I have already been introduced. Golnar, Hubbub’s first employee, handled customer support, hassling shops to pack their orders, administration, and I’m sure a hundred and one things I’ve forgotten. Nicola had been hired to handle communications, marketing, and product development (she still does those things, but now she’s freelance, you should hire her). Then there were Matt and Amy, our drivers. I can’t quite remember the back story, but I do remember neither of them were hired on the basis of their driving experience. The Hubbub philosophy was that our drivers are the face of the company, and we ended up with this amazing bunch of people who would do any customer service team proud.
Hubbub being a small and scrappy startup while everyone technically had a job title and things they were responsible for in practice everyone did a bit of everything. As I worked from home and had a todo list of technology work the length of my arm I tended to stay pretty focused on actually building software, but it wasn’t at all uncommon for Nicola and Golnar to do a day’s work and then jump in a van to do deliveries. “That’s not my job” is a phrase I don’t think I ever heard while working at Hubbub, everyone was willing to step up and do whatever was needed.
My first year with Hubbub is the time when things were most fluid. We still weren’t really sure how to make things work, and we learnt a lot of lessons that built up the foundation of how the company would function. A couple of months in we opened delivery area 2, expanding our deliveries from Islington out into Hackney, and made a big push to acquire more customers. Over the next year we’d open another 3 delivery areas and start to experience the pain involved in having a lot of drivers out on the road at the same time, and the logistics involved in coordinating with a dozen or more shops to make sure they’ve got all their orders ready in time for pickup. I also experienced my first Christmas in the food industry.
Christmas is A Big Deal when you work in food delivery. Typically we delivered as much food in the week leading up to Christmas as we would in a month any other time of the year, which is compounded by problems that would be fine any other time of year being a disaster to customers on Christmas Eve. We once had a customer on the brink of tears because they thought the turkey we’d delivered might not fit in their oven (the driver went in and checked, it was fine). I’d naively booked a couple of weeks holiday over Christmas to make sure I used up my allowance and my wife and I were staying with my parents to catch up with friends and family. I ended up working every day in the lead up to Christmas, mostly moving orders between delivery vans at the last minute to open up a bit more capacity, we’d never needed to do that before so it hadn’t been a priority to build a way to do so.
Growing Up — 2012–2014
2012 was the year we really started growing. With Adam (who had been involved in my interview) joining the company full time as Chief Operating Officer early in the year. His focus was on scaling the company, and particularly acquiring and keeping customers. Adam arriving was really the beginning of Hubbub focusing on being driven by data, rather than gut feel, a theme that would continue throughout its lifetime. It was also the beginning of a period of rapid expansion.
Early in the year we sat down to look at the technology roadmap and get a sense of how long everything was going to take. After a lot of brainstorming feature ideas and some rough estimates we had an answer. 3–4 years. It was time to hire another developer! Sadly none of us really knew how to go about doing that — my hiring had been mostly fluke. Initially we tried traditional job ads and phone screens, but quite quickly gave up on that approach after I’d wasted far too much time talking to completely unqualified candidates. It was time to rethink, and the approach we took was two fold. First, we put in a screening test which I’ve always described as “just high enough for a recruiter to trip over”. That almost immediately filtered out the completely unqualified, but we still weren’t getting any bites from people who we wanted to hire.
Next up was possibly the most cost effective recruiting trick I’ve heard of. The Bacon Bounty. We would offer a year’s free bacon to any developer we hired, and a year’s free bacon to anyone referring a developer we hired. We coupled it with this page, which really only a developer could love.
It paid off when we got a CV from Andy Geers, pitched as a recipe for a great developer. I chatted to him about that the other night, he describes it as “I took my CV and changed the headings”, but when we saw that CV we almost immediately knew we wanted to hire him. The interview didn’t disappoint either, and we offered him the job, only to get some hideous news back. He had an offer from Google. We thought it was all over.
Clearly there’s something a bit special about Hubbub, because after a week or so of umming and ahhing (and being strategically invited out to meet the rest of the team) he declined the big G, and became Hubbub’s second developer. Andy was a perfect fit for Hubbub, passionate about food and helping independents thrive, determined to get the job done, a quick learner, and most importantly someone who cares about everyone he interacts with. After Marisa, I’d argue Andy’s effort to make everyone who joined the company feel welcome has had the greatest impact on Hubbub’s culture. Its certainly something I’ve always admired in him.
With a second developer came some adjustments for me, I had to very quickly learn how to teach someone how things work, both technically and organisationally. Thankfully I had a fantastic student, and within a few weeks we’d massively increased the rate at which we could get new features out. The biggest thing we did with all that extra capacity we automated a huge chunk of Golnar’s job, removing the need for her to track every customer order, special offer, and complaint in the largest spreadsheet I’ve ever seen. That work was the foundation of every special offer we ran from then on, and if you ever got a “we haven’t seen you in a while” email from us that was a side effect of it as well.
It was also around this time that we hired our first fleet manager (although I don’t think that was his title), Carlton McFarlane. He’d started as one of our drivers, but it quickly became apparent he could do far more. He oversaw our ill-fated campaign to personally phone every new customer and tell them about Hubbub (staffed by a few students working after hours from the office), and then he took over running the delivery side of the business, coordinating drivers and vans to make sure customers actually got their orders, further freeing up Golnar to focus on customer service and growing the company.
Christmas 2012 was even bigger than 2011 had been. So big we decided to send out two people in every van — given we still only had the same number of drivers that meant a member of staff in every van, and that meant the first of many photos like this taken from Egham station as I headed into London to meet a delivery van on Christmas Eve.
Christmas deliveries are great fun, and utterly exhausting. You get up long before the crack of dawn, meet a delivery driver, charge round London picking up food, and then drop it off to people who are incredibly happy to see you. In the end its all worthwhile, but I’ll be straight with you, I’m quite looking forward to a quiet Christmas this year!
2013 was a time of changes. Golnar and Nicola both moved onto other things — Golnar joined the civil service fast stream program, and Nicola went to work for Ginger Pig writing about meat. We hired Sonja to replace Golnar (I seem to remember a key thing that got her the job was a great love of stationary, alongside being very good at customer service), and Ben #1 joined us for a few months to work on marketing. Later on that year Hélène joined us to handle the increased customer support we were having to do. In true Hubbub style both Sonja and Hélène were obscenely overqualified, not only being good at the job, but also having done masters degrees in food (sorry, I forget exactly what part of food)!
On the technology side we kept up the rapid pace of development. We launched a redesigned website to try and make it easier for customers to find what they were looking for, entirely overhauled the interface shops used to pack their orders and manage products, and in the course of a fortnight launched not one but two mobile apps.
We’d been talking for a long time about how an iOS app would be a good thing to have but it never made it to the top of the todo list. Every estimate for it was in the order of a month or so and we never had a spare month. At some point Andy and I were chatting, and idly speculated that we could do it in a week if we just had an uninterrupted week without anything else to deal with. Shortly after that conversation we’d arranged to travel with our wives to a cottage in Wales, turn off our email and mobile phones, and build an iOS app. The time between deciding to do that and actually going was spent automating all the things that would interrupt us if they weren’t automated, and as a final flourish we built the first version of a mobile website designed to replace the printed route plans we gave to drivers everyday.
A week later we had an iOS app as well. It was rough, it wasn’t quite ready to go in front of customers, but we had enough that further work on it could be scheduled into day to day life and it would see the light of day. The iOS app never got enough developer time over the years to really thrive, but we did have some fiercely loyal users who would use nothing else.
That summer we moved out of a shared open plan office space into our own space. It was tiny. It was hot in the summer, and cold in the winter. It had an overground line directly outside with a train running past every 3 minutes. It intermittently didn’t have any hot water. But it was ours, and we loved it!
The new office brought with it other changes. Adam and Ben #1 left the company, and that summer I found out I was going to be a father! In amongst all this we were still acquiring customers, still opening new delivery areas, signing up new shops, and hiring more vans and drivers to move orders around. By Christmas we were still doing everything with a team that had now reached the heady heights of 6 people. Marisa, Andy, Sonja, Hélène, Carlton, and myself. By then we may also have hired Sophie, who over the years has been an intern, the first graduate program candidate, and shops liason — she was always endlessly inquisitive, eager to learn, and applying what she learnt to whichever job we’d created for her most recently!
The Big Push — 2014–2016
2014 ushered in another new era for Hubbub. The first few months were focused on getting yet more customers, doing more expansion, and most importantly fundraising! We were yet to make enough money to cover all our costs (although at this point a few areas were starting to make a profit), so Marisa was spending all her time travelling between potential investors pitching why they might like to put some money into the company. Amongst those investors was William Reeve, previously one of the founders of Lovefilm, who since selling to Amazon had been investing in other companies. He was so taken by Hubbub that he decided not only to invest some money, but also to join Marisa in fundraising.
It’s now February 2014, and my life is permanently changed when Arthur George Wood arrives in the world, and go off on paternity leave for (what I thought was going to be) two weeks. This was one of those times when it really sunk home just how different Hubbub was as a company. Arthur’s birth had been very long and complicated, culminating in an emergency cesarean section after 4 days, and my wife Laura being in and out of hospital as a result of complications. We lived in Staines outside London at the time, and just after we came home with Arthur received the only Hubbub delivery to make it outside the M25! I booked another week or so of holiday on top of paternity leave to be around while Laura recovered, and then went back to work while her Mum stayed with us to help out for a week or so.
When Arthur was six weeks old Laura was admitted to a specialist mother and baby unit in Winchester with severe postnatal depression. I spoke to Marisa who true to form didn’t miss a beat when I said I was going to stay with my parents in Southampton for an indeterminate amount of time working odd hours to fit in with visiting hours at the hospital. There was never any question of how we’d make that work, or how long it was going to go on for. The only question was whether we were all ok, and instructions to do whatever needed to be done. In total I don’t think I went into the office between the 16th of February and early May, and not once was there a complaint about it. That was just the Hubbub way — I was far from the only person to experience this.
In the meantime Marisa had closed a £2 million funding round, we were furiously hiring, and William had joined the company as co-CEO. Anytime I mention the idea of a co-CEO people look at me in a slightly shocked way. “How does that work” is inevitably the next question, and thankfully the answer turns out to be “very well”. Marisa and William’s skillsets complemented each other fantastically, with Marisa focusing on the shops and customer communication, while William focused on optimising the operational side of things. William brought with him a push for some much needed process around how we managed the tech team, much needed experience scaling a startup, and a relentless push to get even deeper into gathering and analysing data to determine our direction.
I can’t remember exactly who was around at that time, but we now had about 10 people in the tiny office that had felt a bit cramped with 6, and we had more people joining us soon! We moved across the corridor to another unit in the same office complex, keeping our original space as a meeting/lunch room (and of course getting a proper kitchen fitted in there).
The end of 2014 into 2015 is another period which feels like a bit of a blur to me — I was running on the sleep schedule of a new father, and at the same time it seemed every other week there was someone new! The tech team grew to five people over the year. First on board was Mell, a Scottish frontend developer who so impressed us at interview that we overlooked the fact she’s a vegetarian. I wasn’t around for that interview, but I’m told she came in and basically said we had no idea what we were doing when it came to frontend development. She was right, Andy and I are both backend people really, so we hired her to fix that! Duncan is a trained engineer (an actual one, not a “software engineer” with no real qualifications), and brought the methodical thinking and planning that comes with that sort of training to the tech team. A few months later we hired Jonny, another frontend developer, when we realised just how much work building the new website was turning into. I’m not going to go into everyone we hired for the rest of the company, because I’ll forget someone, but by Christmas 2015 we were once again packed in like sardines. We were a proper company, and we were going to take over the world with our beautiful new website, which had been completely redesigned and rebuilt from scratch.
The Hubbub Christmas party of 2014 will forever be a highlight for me. We’d decided to make the most of our newly fitted kitchen in the meeting room so between us we threw a Christmas dinner with all the trimmings. Goose, roast potatoes, red cabbage, pigs in blankets, the works, complete with party games and drunken chat late into the night. Sitting round that table eating dinner together more than one conversation came round to how much like a family Christmas this was, and it wasn’t just the setting which brought that out, it was how close everyone was to each other.
So much else happened during that time. In that year and a half we expanded deliveries across most of London, and introduced cross docking, where instead of everybody going to every shop drivers would pick up all the orders at once and then meet at a central location (one of the shop’s loading bays!) to swap orders for delivery. That meant we could offer a much wider variety of shops to customers, and effectively run two big delivery areas — North London, and South London. We introduced and later shut down both Saturday deliveries and Hubbub Speedy, a service which would have allowed customers to place an order on the website and receive it 2 hours later. It turned out neither service had enough demand to justify the costs involved in running them, but they still resulted in valuable new systems. Saturday deliveries allowed us to add more flexibility in order deadlines that Christmas, and Speedy led to the development of a new packing system for shops which would move the shops away from printed stacks of paper, and towards using mobile phones and tablets to check and confirm their orders, reducing error rates in the process.
In August of 2015 Laura and I moved to Southampton, away from the ridiculous rents of the London outskirts, and round the corner from all our parents which would give Arthur a chance to see his grandparents, and us a chance to have a break now and again.
Christmas 2015 was huge. So big in fact that we had to completely rework some of our processes to make it possible at all. We stockpiled everything that could keep safely in our meeting room so that drivers would be able to reduce the number of pickups they had to make each day, and cross trained office staff in customer service and delivery procedures. We even hired agency drivers and extra vans because we weren’t going to have space with our existing fleet. We were delivering across London, using a team of professional drivers backed by what I still believe to be one of the best central teams in the industry, and we were giving the big chains a run for their money. Come Christmas Eve we were all exhausted, but also immensely proud of what we’d achieved with such a comparatively tiny team of people. It was also time to start raising money to pay for all these people and the advertising we were doing.
Changes — January-June 2016
By January of 2016 I’d been commuting up to the office in London twice a week, every week, for months. It was taking its toll on me, but with so many people on the team and a tendency for decisions to be made on the fly in ad-hoc discussions it was increasingly difficult for me to stay in the loop and effectively do my job. It also didn’t help that I was exhausted from all the travel. I still loved the company, I loved the job, but it wasn’t working out. I started half-heartedly looking for another job.
For the next few months I continued trying to work out what to do, looking for work, but at the same time not wanting to leave Hubbub. We were still doing exciting work, but my heart wasn’t really in it — this period is probably the low point of my time at Hubbub, but I should make clear that it was still better than any other job I’ve done. The people were still amazing, the company was still doing work I believed in, and with a larger more specialised tech team we could take on ever more exciting projects. In the background Marisa, William, and Roger (our new head of special projects) were working on fund raising.
In May of 2016 I accepted an offer to work as a developer for another startup, Neos. They’re a home insurance company who are trying to improve insurance by including an internet connected security system and fantastic customer service. It wasn’t Hubbub, and I was sad to be going, but I was also going to be able to commute in a lot less. I handed in my notice — I remember taking Marisa to one side and letting her know I’d be leaving was one of the hardest things I’ve had to do in my professional life. I was leaving the best job I’ve ever worked in, and saying goodbye to the best boss I’ve ever encountered. Before I knew it my last day rolled round, and they saw me off in true Hubbub style — a huge dinner of beef wellington for lunch, a leaving card I still have filled with the most lovely messages, and far more drinks than I should have consumed after work. The end of an era. Just not in the way I’d thought.
Interlude — June 2016 to September 2016
I started at Neos, and jumped into the world of insurance policies and door sensors. It wasn’t Hubbub, and it had its annoyances, but it was a good enough job.
Shortly after starting I arranged to pop over to the Hubbub offices after work to catch up with a few people. I usually finished about an hour earlier than Hubbub, but I got an email about 4pm saying they’d just finished a company meeting and everyone was at the pub, so head over whenever I got out of work. On getting there it seemed like something was up, which was to be confirmed when Andy offered to get me a drink. On the way into the pub he told me everyone had just been given notice — at the last minute two big investors had pulled out of the funding round and there wasn’t enough money to keep things going. My heart dropped. Everything we’d worked so hard for destroyed by a twist of fate.
The thing that struck me that evening was how different this was to any other mass redundancy I’d been around. No one was angry or bitter about it, we were just all really sad. We stayed at the pub, reminiscing about Hubbub, lamenting how unfair it all was, and worrying about how Marisa would be taking it. It turned out not to be quite so bad as things looked — there was a tentative plan to keep a small team on and try to keep things running on a smaller scale, but it wasn’t yet clear how (or whether) that would be possible.
The next few months was a barrage of leaving parties, a few of which I managed to get along to (this is the other thing about Hubbub, no one ever really leaves, you’ll continue to see old faces at Hubbub events). At one of them Jin, our designer, mentioned there was a search on to find a single developer to take over and keep things going — Andy was off to work fulltime on his own app, and everyone else had other jobs lined up. The next day I sent Marisa an email:
I miss Hubbub, both the work and the people! Do you think it would it be feasible for me to come back on the basis that I’d be in the office every couple of weeks, and doing lots of video calls in between? The commute is what killed things for me last time round, and with a smaller team I think we could make me being in less frequently work.
Over the next week or two I ended up speaking with Daniela, who’d been around at Hubbub for a couple of years at this point initially doing customer support, and slowly taking on more responsibilities. She was going to be taking on the job of Managing Director, and the day to day running of what was being called Hubbub 2.0. Yes, we all thought it could work, so I sheepishly handed in my notice at Neos only a few months after starting, and started back at Hubbub in September with long enough for Andy to hand everything over.
The Last Roll of the Dice — September 2016 to March 2017
When I was talking about rejoining Hubbub it was made clear that it might not last. This was a final push, everyone involved thought that with everything learned over the last few years and some refinement of the model we could still make it work. We’d moved our sights, no longer aiming to take over the world, we were just aiming to make a profit and become sustainable.
In many ways it was like going back to those first few months all those years ago. We were back to a tiny central team, a few drivers, and a ton of work to be done. Nicola even came back to do comms for us again, having just left M&S to start freelancing!
Work was a rollercoaster. We had a weekly team meeting, and every week it seemed there were new problems to be dealt with, but we seemed to be winning. We merged all of London into one big delivery area, for the first time ever being able to offer all our shops to the whole of London, and giving South London a big boost in sales while we were at it. We pushed further out into the outskirts of London, and we found ways to increase the profit margin on orders. In December Marisa officially went on maternity leave in preparation for her second baby, although true to form she didn’t ever really stop working, even being quite clear that she wasn’t going to call a board meeting short just because she was having contractions!
All things considered, Christmas was surprisingly smooth. It wasn’t as big, or as polished, as previous years, but we pulled the stops out and we made it work. There were a few moments when it looked like it might not, but we made it (even if it did involve William making some of the deliveries in a rented car, and I’m not sure Daniela actually slept at any point during Christmas week).
Come January things were starting to look up, we were turning things around, and team meetings stopped being quite so depressing. Akiba, Marisa’s baby, arrived and cheered us all up a bit more. Daniela had spent the entire period of Hubbub 2.0 pushing for efficiency, and relentlessly focusing in on the tasks which could deliver the most value for the time invested. Considering the size of the team we had we were performing small miracles, continuing to drive down costs and push up profits.
Then it happened. In the second week of March an all hands company meeting was called. The last time that had happened was in June of 2015, and this one was no better. We were out of money, and we were out of time. Our last deliveries would go out that Friday, and then we’d be shutting down the company. My heart sank again, and this time there was no last ditch light at the end of the tunnel for Hubbub. It was the end, 9 years after Marisa first had the idea to make it easier for people to buy food from their local shops.
When we announced the news to shops and customers the response was overwhelming. We were inundated with hundreds of messages of condolence and concern, we’d made a real difference to so many people’s lives, and they didn’t want to see Hubbub disappear. The other overwhelming thing was the number of orders we had over the next few days from people getting one last delivery — this was Christmas 2017, come 9 months early, and with no notice.
For the first week ever, we made a profit.
We proved that with enough customers the model could work, but sadly we hadn’t been able to find enough of those customers wanting to place an order every week to sustain the business, and so it was time to bring it to an end.
The next Monday we all came into the office and starting shutting things down, and packing up. We had lunch together, we joked where we could, and we pulled together to get the job done just like Hubbub always does. Then we went to the pub.
The pub was a celebration of everything great about Hubbub. People from every era of Hubbub came to drink to a crazy idea that we’d all bought into, and to mourn it not quite making it. We were a family, and as families do in times of sadness we all came together to remember the good times.
Hubbub will forever be the benchmark against which I judge all other jobs. From the founder down it was filled with good people, trusted to do their best work, and that’s exactly what we did. Along the way I’ve learnt so much, I’ve grown as a person, and I’ve made some great friends. Little did I know in June of 2011 that innocuous email from Marisa would touch my life in so many ways.