# S01E06 · Transcript — Cilliers Geldenhuys Full transcript of [The Useful Podcast](https://useful.ventures/podcast) Episode 6, a founder-to-founder conversation between host Gokul Bala and Cilliers Geldenhuys, co-founder of [AllWage](https://www.allwage.com). Back to the [episode page](https://useful.ventures/podcast/cilliers-geldenhuys-allwage). This transcript has been lightly edited for readability. Filler words and false starts have been removed and the meaning has been preserved. Any errors are mine, not Cilliers's. ## Meeting in a university residence **Gokul** I love this detail about AllWage. You, Luke, and Johan all met at university, living in the same residence — two computer engineers and an industrial engineer. Were you friends back then? How did the thing blossom? **Cilliers** It's an interesting story. We met in res and had some subjects together, and became friends there too. After res we went in different directions, and the project actually brought us back together. It was much easier to start from nothing because we were already friends and had spent time together studying. There are still memories of sitting on Luke's bed while he showed us how to do calculations for certain subjects. Some fun moments. **Gokul** Did you guys talk about working on a business back then, or was that out of the blue? **Cilliers** Maybe five minutes were spent on that discussion in res — nobody really thought about the future then, it was more about acing a subject and having a few beers on the weekend. After we started working, the discussions became more frequent. That's when it really started, after res. **Gokul** You all worked on mining projects after graduation. What did that look like? **Cilliers** Luke and Johan went to a software company and worked together there — Johan also spent a year in Japan before coming back. I got a job at a mining contractor, and saw an opportunity with the leadership there to monitor machines better and improve efficiencies with software. I started speaking to developer friends, and through those discussions Johan and Luke became part of it. I contracted them in while I was still permanent at the mining company. Eventually we decided we could build cool things together, so we took a stab at it and went all in — that's how the three of us started Agrigistics, which we rebranded at the start of this year to AllWage. ## From monitoring machines to monitoring people **Gokul** How did you transition from monitoring machines to monitoring people? **Cilliers** One of our friends is in farming, and he said one of his clients was having trouble monitoring farm workers. We thought, we're already monitoring machines, maybe it's basically the same thing — which is not true, but that's how we went all in on the problem. We started monitoring what farm workers harvest, what they pack, the diesel and chemicals and fertilizers they use — basically everything a farm worker touches. That became quite wide, and we struggled to scale because every farmer has a different setup and we ended up with too many modules. We realized the main problem was actually time and attendance — the flexibility you need around wage workers to manage them properly. So we pivoted into that direction. **Gokul** How big a jump is that, from monitoring wage workers to becoming a payroll company? **Cilliers** It was natural. On a farm you also want to see the hours alongside what a worker harvested, and that's how time and attendance started. But the natural next step is payroll, because those hours have to go somewhere. So we now have four main legs: time and attendance, with flexible hardware and software options to clock people on the ground; payroll, live and helping wage employers; WhatsApp-based communication with employees; and HR. Our main focus now is helping employers manage, pay, and grow their wage workforce more easily. **Gokul** The original idea was almost beautiful — tracking who participated in harvesting and packing so they could be incentivized better. That's not really surveillance. How did you land on that framing? **Cilliers** We didn't have startup experience going in. We got to our first farmer, he told us all his problems, we talked to another farmer close by, and I'd say the mistake we made early on was listening to everything our early clients told us. We wanted to solve all the problems, and we knew we had to talk to customers, but we didn't have a good filter on how to weigh that against where we really wanted the product to go. So the first ten clients pulled us in those directions — not a bad thing, because it showed us what we needed, but also a bit of a mistake. **Gokul** What was it like deciding to leave your jobs and pursue this full time, all three of you? **Cilliers** I always knew I wanted to try building something, I just didn't know how to start. Once we started working together, those conversations happened more and more. Then Luke and Johan phoned me one morning and said they'd left their jobs and were going full out on this. I still needed to work, because we had no clients and no money to pay anyone — so they said they'd live off bread and milk while we tried to get the first few clients. It took about a year before we could actually pay them, and for me to move over too. **Gokul** How scary was that? **Cilliers** It was scary. It helped a lot that we did it early in our careers — I think I was around twenty-six, no major responsibilities, no children, I'd just gotten married. I knew that if it didn't work out, I still had a degree to fall back on, though we tried not to dwell on that and just focus on giving it two years and then deciding. We checked in with each other regularly, and as we kept growing we kept telling one another it made sense to keep going. ## Ghost workers **Gokul** I'm really curious about the problem of ghost workers — I don't think anyone listening has heard that term. Can you explain what it is and give us the story of the problem it causes? **Cilliers** The term resonates a lot with countries that have a lot of wage workers, often uneducated and lower-paid. Businesses have sites in a lot of different, very rural areas, sometimes with no internet or power. Someone will try to get a friend into the company by adding them to the time-attendance register, and because the office is so disconnected, they just see there's a new name and assume the site manager added someone who's already working. That person lands on payroll, and months later, if management audits the sites, they discover twenty ghost workers, or whatever the number is. We try to solve the disconnect between the office and all these rural sites. **Gokul** Is this more like wage theft, or impersonation? **Cilliers** You can call it wage theft — we just call it ghost workers. That worker doesn't exist. There's a related but different problem where people try to steal hours: they clock in early or clock out late to steal overtime if there isn't a proper system in place. It's usually a small share of the workforce — most wage workers genuinely want to work, earn their salary, and be loyal. But you need a flexible system to manage the smaller group that tries to manipulate it, so you can grow the people who actually want to be there. ## Building for the ground: WhatsApp, RFID, and facial recognition **Gokul** Your users are wage workers, and you've built clocking-in through WhatsApp, RFID wristbands, biometrics, and facial recognition. Why that mix of tools, and why WhatsApp? **Cilliers** We started with RFID bands, from our farming direction — they clock activities quickly, but the main problem is people can exchange bands. So we added facial recognition too, and some clients, especially in agriculture, use both: facial scanning when workers enter and exit, and bands to monitor activity within the blocks where they harvest. WhatsApp helps smaller teams, especially in construction — a crew of three driving between sites usually always has WhatsApp, so they clock in there without needing hardware, and we do recognition by their cell phone number, along with GPS location so employers can set geofences and confirm clocks are happening in the right places. We also give employers the option to run facial scanning through an app on their own device, so a foreman running a team of thirty doesn't need to invest in hardware. And for manufacturing facilities we support a fixed facial scanner linked to a turnstile for proper access control. We try to identify what each industry actually needs on the ground to get the data collection right. **Gokul** How much of the design was shaped by the environment — weather, connectivity — versus the type of worker? **Cilliers** The environment has a big effect. On very rural farms with no signal, offline clocking capability was the first thing we needed — a device clocks offline and syncs automatically once it reaches Wi-Fi or cell signal. Different worker types matter too: farmers, for example, love the facial scanning with armbands because they want to monitor harvesting block by block, and they don't tend to use the WhatsApp option. **Gokul** Do you have Silicon Valley competitors who don't understand the user type, or what's the landscape like? **Cilliers** We have competitors, but I wouldn't call them Silicon Valley — there's a handful working in the same space we are. Being in a third-world environment is actually good for us, because our focus is going forward into other third-world countries, and I think most AI development is currently focused on first-world problems, so there's a lag in the developing world that feels like a safe space to operate in. We actually know a few competitors and go for coffee sometimes — keep your enemies close. **Gokul** Since you're building facial recognition and geofenced tracking, and pointing it at people with relatively little power in the situation, how do you think about the social responsibility that comes with that? **Cilliers** We started building for the employer and still mostly do, but as the ecosystem evolved we started focusing on employees too, because if they benefit, the employer benefits, and so do we. We started feeling the responsibility to help employees get into the right company so they can grow and stay, and so businesses can think about expanding the workforce instead of shrinking it because it's too hard to manage. That employee side is going to expand a lot in the coming years — HR is one direction, and financial benefits like early wage access is another we're looking at. ## Loan sharks and early wage access **Gokul** Can you explain the loan shark problem for wage workers? **Cilliers** Minimum wage in South Africa is quite low — someone might make around five thousand rand a month, which isn't much once you account for family and other needs. People end up borrowing from loan sharks at very high interest rates, and when they can't pay it back they sometimes run away, and the employer loses the worker. Companies are addressing this with early wage access: employees can access wages they've already earned at very low rates instead of borrowing from a loan shark, so the employer isn't worried about people disappearing, and the employee gets access to funds without predatory interest. It's a massive market, and it's compounded by online gambling, which disproportionately affects the low-income workforce — so you have that vicious circle of low income, gambling, and loan sharks waiting. **Gokul** You started as a tool for employers and now want to serve employees too. Does that change what the company is? **Cilliers** I wouldn't say it changes the company — we've always said that if we build for employees, the employer has to benefit too, because that's where we started. We want businesses to grow so the economy thrives. That keeps us on the straight and narrow, because building only for employees without that constraint would get too wide and hard to scale. ## Staying bootstrapped **Gokul** You've built the business starting in your own free time, taking very little money, and are now profitable with no big raise. Was that by choice, or how the cards played out? **Cilliers** Definitely some luck. I was traveling with my wife and met a guy at a backpackers' in the Eastern Cape — we had a long conversation, he shared books with me, and I took his number. I told Luke and Johan about him, and we started reading the books he'd given us. It turned out he was a fairly big figure in the South African startup space. He saw that we had high agency to get things done, and became one of our angel investors, helping us a lot with when to raise, when not to, and how to grow. That was the massive luck point. From our side, the high agency was the biggest thing we did right. **Gokul** Did any investor give you something more valuable than the cash? **Cilliers** Yes — and still does. He meets with us almost every week for an hour. We're almost like friends at this point; it's a very open, honest, sometimes judgmental conversation, and it's helped massively up until today. **Gokul** Did being profitable change how you thought about moving forward, compared to a funded business? **Cilliers** Because we didn't raise much capital early on, I was anxious every afternoon — you don't have a lot of burn room. Once we became profitable, it got much easier to think about longer-term strategy across every function of the business, rather than just about the next client to sign and retain. **Gokul** Would you consider raising now that you're profitable? **Cilliers** Not at this stage. We keep growing faster year on year, month on month. We have problems to solve, but we have a good plan and strategy, and more cash would help without being what's holding us back. We think we can build this to be massive while keeping most of the value of the company rather than giving it away. We're also wary of a large VC coming in and influencing our strategy in ways we might not choose ourselves. Currently we have full control over strategy, and that's amazing for us. ## Radical transparency **Gokul** You practice radical transparency with your team — they see the full financials and the strategy. Why do you do this, and what inspired it? **Cilliers** You've done good research — we also got this from our angel investor, Philip. He told us transparency helps from the start, because the team can see the growth and rally around one number the whole company is working toward. So weekly or biweekly, we go through where we are as a company, show all the numbers, and the team knows exactly what's working, what isn't, and why we're focusing on different things. **Gokul** Has it ever backfired, or changed how your team works? **Cilliers** I wouldn't say it's ever backfired. There's a risk that someone leaves and shares the financials elsewhere, but I don't see that as a massive risk — maybe just not a nice feeling. It's never caused an issue for us. ## Winning South Africa, one industry at a time **Gokul** You've said what mattered most to you wasn't headcount, but category dominance on wage workers in South Africa. What's the strategy to get there? **Cilliers** First, we try to keep the team small with people who have high agency and really want to make a difference on this problem — which AI makes more possible than it was a year or two ago. We try to leverage AI across the business and move people around as functions get automated or improved. We also went through a phase of reading Crossing the Chasm. It made us realize that to win South Africa as a market, we needed to focus down into one industry — construction — win that first, and then replicate the knowledge into manufacturing, agriculture, and the rest. So our marketing became construction-focused: we film on construction sites with construction clients, and talk about construction-specific problems. We still get leads from every industry, but the focus shifted from agriculture in the past to construction now. The next phase was landing a marquee enterprise client in that industry — which we've done — and using that to show medium and smaller businesses in the same space how we've helped companies like theirs, because they tend to want to follow. **Gokul** Your ideal customer is a company with twenty-five to two thousand wage workers, though your sweet spot is more like twenty-five to six hundred. Why is that the gap everyone else ignores, and what does enterprise software get wrong about them? **Cilliers** A lot of our competitors focus on the big brands with twelve to twenty thousand-plus employees, and we feel they're missing out on small to medium businesses. There are real disadvantages there too — sometimes no structure, no HR person, a one-man band — but we've built to serve that section of wage-worker businesses specifically. **Gokul** How do you actually get a construction company on board, given the industry runs so heavily on pen and paper? **Cilliers** About half the industry is still on pen and paper. We usually start by mapping how they currently manage sign-in and sign-out, which is inherently inaccurate because anyone can sign in and out at any time — nobody's really checking those sheets. We bring in the right stakeholders, set up our system to mirror their existing process but make it accurate, and then show the value: now you can see the clockings at nine in the morning, see which three people on a team didn't clock in or clocked in late, and start real conversations with the manager. We've developed a consulting layer over the years where we walk clients through good practices as we onboard them and show them where they'll actually see value. **Gokul** Where else do you see AllWage expanding — you've mentioned Namibia and Mozambique? **Cilliers** We actually stopped selling into two countries outside South Africa because we want to win South Africa first before losing focus elsewhere — and the market here is big enough given how many industries we serve. But almost any country with a lot of businesses and a lot of wage workers is a candidate: Brazil, Spain, Mexico, most of Africa. Kenya and Nigeria are massive markets too — I saw a stat that Nigeria's construction sector may be growing even faster than ours. Right now we're a hundred percent focused on South Africa. **Gokul** Six years in, profitable, around fifty thousand wage workers — did you think this was possible? **Cilliers** I definitely didn't think we'd end up building payroll for wage workers — that wasn't the plan when we started. The problem brought us here. Looking back, nothing feels that different in practice, even with bigger numbers now than when we had zero in the bank. I think that's a good thing, and one of the better practices between the three of us. **Gokul** What does winning look like for you? **Cilliers** Right now, winning means winning the African market for wage workers — that's the first big phase, and we're not thinking much past it. Having a good, sustainable team at that stage, actually seeing the difference we're making in these companies and getting feedback that what we're building is having real impact — that would be winning. ## AI, activation, and the hardest part of the job **Gokul** You've said AI adoption isn't really an engineering problem anymore, it's a user-adoption issue. What made you realize that? **Cilliers** We meet every three months with other founders in Cape Town for dinner, and everyone talks about their stories — AI is the big topic now. The engineers there kept saying you can ship very fast now, but you still have to understand what to build, talk to clients, and get feedback on what you've shipped. You can't skip that. For us, in B2B in a developing country, it's especially important to keep talking to customers, because adoption of something new — like WhatsApp clocking — takes real feedback. Engineers can build very quickly now, and we try to make everyone on the team a builder, but you still have to close the loop on whether it actually solves the problem and what value it creates. **Gokul** How do you use AI internally, and is it in your product? **Cilliers** We started by building it internally first — marketing, sales, onboarding, engineering, product — which let us scale with a smaller team and get more efficient. Where a job function got replaced, we moved that person somewhere else rather than let anyone go. Then we started adding AI into the product itself, which is the stage we're in now. Because we handle a lot of sensitive information, we move carefully: talk to clients, figure out the one thing to build really well, build it, review it with them, and figure out what's next. Most of our time right now goes into those client conversations rather than building. **Gokul** What's been the hardest part behind the scenes over the last few years? **Cilliers** Saying no. Everything feels important when you're working across a lot of functions and stakeholders, and figuring out the next thing to build for maximum leverage is genuinely difficult. Right now our main struggle is activation — getting a client from disconnected, often non-digital, pen-and-paper information onto a live system efficiently, with all the right stakeholders on board. That's the problem we're trying to solve now, to move faster. ## Building the team **Gokul** What's gone into building the team? **Cilliers** It's been difficult at times. We try to hire people who manage themselves and have high agency — I remember an article comparing employees to bullets and the casing that holds the bullet, about the type of person who can handle a lot of different functions rather than being so specialized they can't be used anywhere else. That's still our focus. What we struggle with now is finding really good engineers who use AI heavily and are at the forefront of it — a lot of engineers, and people in other roles too, just haven't made that shift or aren't curious enough. **Gokul** I think AI adoption is at an all-time low relative to what it could be, even among people who are using it. **Cilliers** Definitely. I'm not sure why. For me it happened because I was curious about a specific tool, used it, and it solved a problem — that made me a believer, and then I wanted more and more. If you're not a believer and don't use it correctly, I think the gap between believers and non-believers is only going to keep widening. ## What winning looks like, and family **Gokul** You've spent your career building software for people who don't usually get anything built for them. Was there a bigger focus running through your career? **Cilliers** I've talked to my parents about this — I'm not sure exactly where it started, but from a young age I'd say things like "if this was my business, I'd change this" when we walked into a restaurant. I told people at res, half-joking, that one day they'd work with me. That narrative kept running through my childhood, even though I didn't actively work toward it after university — I enjoyed what I was doing at the mining company. But once the opportunity came to actually do something, I didn't think twice. **Gokul** When you decide to retire, what do you want to be known for? **Cilliers** I want the flexibility to spend time with my children and my wife, traveling and doing the things we want to do throughout the year. But I also want to know I reached that while doing right by the businesses and employees we've worked with. From the start we wanted to build something great while helping people — we didn't know what the end result would look like, and I'm not sure we're fully there yet, but we always wanted to make the world a bit better. **Gokul** How does your family factor into the business day to day? **Cilliers** We're growing our family, and our whole company is remote, so everyone works from wherever they want — that already gives us more flexibility to spend time together. But building a company still takes a lot of time during the week. At AllWage specifically, we try to keep hours limited to eight to five, and only occasionally work nights or weekends. That's worked well for us — we expect a lot from people during working hours, but we try not to force anyone outside of them. --- Back to the [episode page](https://useful.ventures/podcast/cilliers-geldenhuys-allwage) · [AllWage](https://www.allwage.com) · [[podcast|All episodes]]