Building a Global/Local Platform with Etsy

In a recent and very informative fireside chat, Linda Kozlowski, COO of Etsy was in conversation with Sarah Moran, CEO of Girl Geek Academy. Key themes of the discussion included the challenges in building a “global/local” platform, making sure you are addressing the right audience needs, and in a two-sided market place, knowing how to balance the interests of sellers and buyers.

etsyOrganised by StartupVic and hosted by inspire9, it was yet another example of how fortunate Melbourne is to attract and host so many leading global figures in the startup world, willing to share their insights (as well as learn more about the local startup scene – which probably does not get as big a rap as it should, especially in mainstream media).

With previous operational roles at both Alibaba and Evernote, Linda brings a strong combination of experience in tech and market places, and describes her current position as covering “CX and revenue from end-to-end”. Her primary focus is on product development, global expansion and addressing sellers’ problems.

At Etsy, the aim is to develop a global product platform that is culturally diverse. There is a natural tension between fully localised customisation (which can be costly to maintain – translation, version control, managing updates), and a “best of breed” model that can serve most users (which can lead to too many compromises). So instead, Etsy pursues a strategy of locally originated products and features, combined with open APIs.

Etsy has also identified sellers as their core audience. This means that so much of the CX is actually determined by sellers’ needs, to whom Etsy then serves up customers to the platform via social media, content marketing and SOE. Etsy sees this as a point of differentiation when considering traditional retailers who often end up squeezing their suppliers on pricing and margins.

In balancing the needs of two-sided markets, again Etsy focuses on the seller first – because buyers want to see depth of inventory and a range of quality products, so get sellers on-board and the buyers will come.

Asked how Etsy avoids buyers going direct to suppliers, Ms Kozlowski commented that it all really depends on where transactions happen: make the CX is so sticky that sellers want to stay on the platform, offer great seller features, and bring in quality buyers. On the other hand, it’s a healthy market place, so there is no mandating or exclusivity – because of course, successful sellers will sell in multiple places.

Etsy believes in investing in the right technology and the right marketing at the same time. For example, although the business was started in 2004, Etsy only began brand marketing in 2016. The types of marketing deployed include storytelling, identifying key differentiators, understanding customer influences, extensive content strategy, plus performance marketing (site traffic, SEO) to see what else customers may be looking for. According to Ms Kozlowski, a retail site should spend about 20-30% of revenue on marketing, otherwise you are buying traffic and customers (red flag to investors!) Alternatively, establish some ROI goals for marketing costs, especially performance marketing, and during the startup phase (up to 1 year) begin with 15-20% of revenue.

At the heart of Etsy’s business model is an evolving technology and e-commerce platform that allows micro-businesses to run at scale. The USP is a community of network effects: 1.7m sellers, 27m buyers, 40m products, 12,000 “teams”, whose “captains” recruit other sellers. (Australia is actually in Etsy’s Top 5 markets.) In addition, Etsy aims to act with integrity in respect to dispute resolution and addressing fraud – they use machine learning to detect rogue sellers (although often, buyer disputes are a reflection of seller inexperience, rather than fraudulent behaviour). Etsy also offers training and support to resolve disputes, and the community is very good at policing itself.

What are some of the unexpected challenges of market places? When it comes to distinguishing between commodity items and unique creative products, technology will under-price and displace commodity goods that are easily made. It’s also important to build human-to-human connections, and to have a global perspective – for reasons of quality and diversity, and not self-limiting your business, especially when it comes to managing different market and economic cycles. Remember to “follow the data”, and anticipate demand and trends. Among some of the technology challenges, dealing with different devices in different countries can be an issue.

In conclusion, Ms Kozlowski offered some advice for anyone thinking of launching a market place:

First, consider why there are very few local e-commerce markets in Australia (from my personal perspective it’s a complex mix of retailers getting burned in the dotcom boom/bust, cosy market duopolies, and perpetual geo-blocking…). But let’s not forget that Alibaba has just opened its Australian & New Zealand HQ in Melbourne, likely to ruffle a few feathers in the retail sector.

Second, embrace a local/global mix (for the reasons mentioned above)

Third, don’t price too low (although Australia is notorious for being a “price-sensitive” market…).

Finally, the conversation ended with a tantalising glimpse of “the future consumer” (Gen Z) where makers connect directly with buyers.

Next week: Spaceship launches the future of superannuation

Gigster is coming to town….

Melbourne’s Work Club recently hosted Gigster Senior Project Engineer, Catherine Waggoner, in conversation with Venture-Store’s George Tomeski. Part of Startup Victoria‘s Fireside Chats, it likely herald’s Gigster opening an office in Melbourne, to service local clients and to tap into the local developer community.

gigsterFor the uninitiated, Gigster describes itself as the “world’s engineering firm”, that helps clients scope, design and build software, apps and digital products. Using an established product development methodology, and drawing on the resources of a 1,000 strong network of freelance designers, developers and product managers, Gigster is taking much of the pain out of the costing and requirements process for new projects, as well as building a growing client base of enterprise customers.

Not mincing her words, Ms Waggoner opened her remarks by commenting, “The software development industry model is f*#$ed”, because:

  • Requirements are poorly defined
  • Scoping is laborious
  • Development costs blow out, and
  • The whole process is not very transparent and not very accessible.

As a case in point, she mentioned the significant cost disparity between what some digital design agencies or app studios might quote for building an iOS product compared to what Gigster would estimate. By: breaking projects down into the distinct stages of scoping, design and pre- and post-MVP; only engaging the “best of the best talent”; using proprietary tools both to estimate fixed rate costs (rather than billable hours) and to define and source solutions; and re-using content from a library of “Community Software” resources, Gigster is able to deliver quality projects in shorter time, and on more modest budgets. For example, based on the large number of projects that they have fulfilled, their “Gigulator” estimating tool incorporates 5,000 possible features.

From an investor perspective, Mr Tomeski mentioned that the “VC inflexion point is getting much earlier” in tech startups. Meaning, with lower development costs (and potentially, reduced valuation multiples), investors are looking to get in sooner, with lower exposure, but still generate reasonable returns on exit, thanks to cheaper establishment costs.

Of course, Gigster sits at the heart of the gig economy, a huge issue when it comes to discussing the Future of Work. Interestingly, many of Gigster’s contractors are themselves startup founders, who freelance while building their own businesses. But such is the strength of the network, something like 35%40% of their contractors work full-time for Gigster – they like the flexibility combined with the continuity. Many of the contractors are referrals from existing team members, and a number of teams (known at Gigster as “houses” – presumably a frat thing?) have bonded to such an extent that they get allocated specific projects to work on together, even though they themselves may be working in different locations, based on previous projects.

Working for Gigster is probably a career choice for some contractors, because there is a variety of projects to work on, and the opportunity to be involved from start to finish. Which may be the opposite if working in a more corporate or enterprise environment, where work may be routine, repetitive and reasonably narrow in scope.

If Gigster does decide to set up shop in Melbourne (with encouragement from
InvestVictoria) they will be joining the likes of Slack, Stripe and Square, tempted by financial and other incentives. Such a move may challenge a number of local digital agencies, who will face even more competition for talent and customers.

According to Ms Waggoner, enterprise clients represent 40% of the business, and should comprise 60%-80% very soon. Not only that, but the average deal was initially $15k, now it’s more like $100k. However, enterprise clients have a much longer sales cycle. Plus, many innovation teams within enterprises are more like loosely formed groups of niche experts, so they need training on how to think like a startup. When you consider the greater dependency on legacy software by corporate clients (where it may make financial sense to retire some assets and build afresh, but the emotional disruption can be huge…), combined with the greater emphasis placed on after-sales service, Gigster has had to adapt its business model accordingly.

But Gigster must be doing something right. They’ve stopped outbound marketing and prospecting, relying on in-bound leads, repeat business and client referrals. There has been a shift from a sales focus to a customer focus, complete with a dedicated customer success team.

A number of audience questions related to getting VCs interested in your idea: What do they look for? How do they assess opportunities? How far should you go in building a product before you can attract funding? What’s the best way to validate an idea? etc. Much of this is about product/market fit, building the right team, getting customer traction, and executing on your strategy (aka Product Development 101.) As part of her closing comments, Ms Waggoner noted that unlike some of the high-profile VC funds (e.g, Y-Combinator, Techstars and 500 Startups) many VCs are becoming more sector specific, because they prefer to invest in what they know and understand.

Next week: Building a Global/Local Platform with Etsy

The latest installment of Startup Victoria #pitch night

The numbers were out in force for the August edition of Startup Victoria‘s monthly pitch night. A full house (no doubt helped by a new beverage sponsor…) heard from another batch of startup hopefuls, operating in very different sectors: medtech, recruitment, food logistics and domestic services. Despite some AV issues, this event showcased some interesting businesses, all of them demonstrating some impressive early stage traction.

In order of appearance, the night’s pitches came from:

VideoMyJob

Launched in April 2016, this online tool allows recruiters and hiring managers to film, edit and share their job ads. The business already boasts more than 60 clients (some of them very high-profile), with the data suggesting an 82% higher success rate in hiring outcomes. This performance is largely attributed to the simple fact that candidates spend up to 4 minutes watching a video ad, rather than the average 12 seconds candidates spend reading a text-based ad before they submit an application.

The tool, which runs on a mobile device, includes a tele-prompt feature, in-app editing functions, a one-step process to publish to social, plus e-mail. Customer pricing is based on a $79 monthly subscription to place unlimited video ads. One reported benefit for clients is much stronger candidate short lists.

Given the changing dynamics in the recruitment market, where companies are finding themselves competing for talent and striving to become employers of choice, any new hiring solution has the potential to be a game-changer. Which is what the founders are probably banking on as their exit strategy, with a likely trade sale to a complementary recruitment platform.

PredictBGL

This medtech startup (previously known as ManageBGL) offers an app-based solution to help diabetes patients manage, monitor and predict their blood glucose levels. Despite regular patient testing, according to the founders, 80% of the data is actually ignored.

Able to offer more “real-time” testing, the app claims to fix wrong insulin doses within 3 hours (not the usual 14 days with traditional clinic-based testing), offers more precision dosing, and predicts patient levels up to 8 hours ahead.

It also has the option to incorporate live exercise data (from wearables), and serve patients who can’t afford expensive insulin pumps. As well as paying a monthly subscription, patients are also paying for insights based on the data. With a $10 per month fee, over 80% customer retention rates, and around 600 sign-ups per month, the app is breaking into the US market.

Asked about potential risk factors and the margin for error in patient testing, the founders explained that the user results are somewhat conservative, so they are embarking on clinical trials to refine the analytics.

Jarvis

Billed as “your very own personal butler”, Jarvis is one of a number personal concierge services, catering to the time-poor, inner-city residents who want to outsource domestic chores and errands.

From $33 per week (and an average of $55), Jarvis differentiates itself by offering a more personal touch, because the business hires and trains employees, rather than using freelancers or contractors.

Launched in January 2016, Jarvis is experiencing 20% growth per week, 90% customer retention, high referral rates and generating 10-15% margins. The founders are working on their logistical efficiency – routing, grouping – and deploying scalable technology – such as cluster algorithms. Pat of the attraction for clients is the fact that Jarvis does not see itself as a transactional service like some freelance and task-based apps and platforms.

The panel of judges asked about the risk of being disintermediated (by their own employees going direct to client). Jarvis claims that their key defense is the proprietary Butler app for employees.

Pantreeco

Last up was Pantreeco, which was established in 2014, with the goal of building “productive partnerships in food” by streamlining the logistics and supply chain communications between food suppliers and buyers.

A self-styled “co-commerce” solution, Pantreeco includes a messaging tool between producers, wholesalers, distributors, restaurants, cafes, grocers and providores.

Offering a freemium SaaS model (based on a per customer per channel basis plus commission), Pantreeco is in the process of taking its model to overseas markets via some major international expansion.

Asked by the judges about the competition, such as TradeGecko and Unleashed, the founders stress that they are not simply an e-commerce or inventory management solution. Instead, Pantreeco developing a range of integration services in response to customer demand – e.g., invoicing, accounting, communications as well as inventory management with 3rd party platforms such as Xero, ZenDesk and SalesForce. They also have plans to on-board major enterprise clients in the food and beverage industry.

Based on the audience voting, Pantreeco took out the honours on the night.

Next week: When robots say “Humans do not compute…”

Customer service revisited: Navigating The Last Mile

From time to time, I like to comment on the current state of customer service, because this is still one of the key areas where companies can differentiate themselves. So, based on recent experiences with a bank, an insurer, a telco and an e-commerce site, I’m sharing my thoughts on the Last Mile – where even great products and great companies can fall down due to their inability to truly understand the customer experience they create.

Image sourced from LinkedIn

Image sourced from LinkedIn

1. The Bank

After waiting over 30 minutes in a call-centre queue, I eventually spoke to someone who said she could help me with a query regarding the disparity in the amount and rate of interest earned on one of my savings accounts. But first, I was given a choice: either accept an instant $50 “goodwill” payment now, or wait for the outcome of her investigation. Because the amount I was querying was several times that offer, I requested she look into the matter further.

Leaving aside the fact that she failed to get back to me within her stated timeframe (I only managed to re-engage the bank when I queried the lack of response via their social media account…), it transpires that she gave me incorrect product information. This underscores one of my main complaints about customer service – inadequate product and process training. Her supervisor who picked up the query then offered me a $10 “goodwill” payment for my trouble (overlooking I had already been offered $50!).

It was only when I insisted that the amount I was potentially out-of-pocket was closer to $300, and following a protracted and somewhat terse negotiation did the supervisor choose to exercise her (undefined) discretion and settle for an amount in between $50 and $300. While the outcome was closer to what I had expected, the customer service process and experience were far from satisfactory.

2. The Insurer

My home and contents policy recently came up for renewal. I noticed that, even with a customer loyalty discount, the premium increase was far higher than current CPI. It seemed to me that a previous “special discount” I had been offered when I last updated my policy at a bricks and mortar branch, rather than by phone or online, was now being clawed back (and then some) with the latest premium increase.

So, I shopped around online and found a better deal. When I rang the original insurer to advise them I was cancelling and taking my business elsewhere, they said: “Is there anything we can do to keep your business?”. My response was, “Too late.”

I accept that premiums may have to increase. But rather than simply sending out a renewal notice asking for more money, I think the better strategy would be to provide an explanation for the increase, and demonstrate the additional value I would be getting for renewing my policy. I resent being taken for granted, because the insurer clearly assumed I would simply pay the increase on demand, and only attempted to offer a better deal when I rang up to cancel.

3. The Telco

Late last year, I switched telcos, because the service was increasingly reliable, and I had experienced poor customer service from the start of my contract. In the process of transferring my mobile, fixed line and internet accounts, I notified the telco that I was dissatisfied with their service, and was taking my business elsewhere. I also initiated the return of my telco-supplied modem, to avoid incurring any additional fees or expenses. 

However, the telco continued charging me for certain services, long after I had discontinued using them, and 2-3 months after they had been ported over to my new service provider.* I requested the refund of the overpayments. The telco refused, because they claimed they had not actually been formally notified that I wished to cancel the services. So I lodged a complaint via the TIO, but the telco still denied any liability, and refused to refund my money.

Eventually, a TIO Investigation Officer was assigned to my case, and he agreed that on any reasonable reading of my complaint, the telco should have concluded that I was cancelling the service. The telco continued to resist my request for a refund:

E-mail received May 31: “[We have] reviewed the complaint and have decided that we will not be changing our position on the matter.”

I believe that the Case Officer then suggested that the telco listen again to the calls I had made, and place them in the context of the other contemporaneous events and the full history of my contract. He also advised the telco that he was prepared to initiate a full and formal investigation of the complaint.

Only then (and in a remarkably speedy U-turn, worthy of a politician) did the telco respond:

E-mail received June 7: “Thank you for your time and patience throughout this case, it is really appreciated (sic). We apologise for the poor level of service you’ve received that led you to escalate to this point. This is not the kind of service we want our customers to experience and it’s very unfortunate that you have to go through this, especially after you cancelled as a result of the poor service.
 
We will be crediting the account with $XX for the period from the XXth December 2015 to the XXth February 2016 when the service was active after it should have been terminated.”

I’m clearly grateful to the TIO for their assistance, but frankly, it shouldn’t have to get to that point. For an organisation that prides itself on superior customer service, the telco in question clearly does not understand customer experience.

4. E-commerce

There are several reasons why I prefer to order online, rather than buy from local shops: convenience, choice, availability, service and often price as well. Speed of delivery is usually not a factor, especially when ordering from overseas (although in many cases, ordering from overseas can be quicker than buying from a local online store).

However, I’ve recently experienced some delays in overseas deliveries, and upon investigating the matter, discovered that, quite apart from a lack of knowledge on the part of some customer service reps (that old chestnut), the multiple links in the supply chain can result in mis-communication and mis-alignment of their respective operating systems.

For example, if the online retailer does not actually fulfill the order, or if they or their nominated carrier outsources customs clearance and/or the final delivery, there may be as many as 6 or 7 hand-off stages in the process. Unless all the back-end platforms talk to each other (and in the same language), the risk of stuff falling between the cracks is very high.  (The notion of same-day delivery by drone is probably some way off…)

What is particularly frustrating is when one part of the vendor’s website has the (overdue) ETA as one date, but another part of the same website shows a much later ETA – even within a single platform! Perhaps if retailers got their upstream systems in order, the Last Mile would be more likely to take care of itself?

*Footnote: My original provider is merely a re-seller, and therefore is subject to wholesale access provisions. According to some information I received from my new provider, it is illegal for a telco to charge for services over which they no longer have any control or access.

Next week: Field report from Melbourne #Startup Week