Let me preface this by saying that I truly believe that at least half of success in business can be attributed to luck, with the other half being execution. Although it is fruitless to try and document something as idiosyncratic as luck, I will discuss what I believe were the key principles that allowed us to scale our business to where it is today.

As a quick primer, Althea.kr is a Korean Beauty ecommerce business. We were founded in Malaysia in July 2015 and in the past year have expanded our operations to Singapore, Philippines, Indonesia, and Thailand. Within one year, we were recognized by Forbes as the largest Korean Beauty website in Southeast Asia with an annualized revenue run rate of $10 million.

Luck aside, I think we can attribute our ability to grow so quickly to three principles – focus, localization, and financial discipline.

1. FOCUS

What differentiates Althea from many ecommerce businesses is that we exclusively focus on one niche – Korean Beauty (K-Beauty). In the past 10 years, the K-Beauty industry has experienced CAGR of 33.8% and a significant amount of that demand comes from Southeast Asia.

We estimate that the addressable market for K-Beauty in the region alone will be $8 billion by 2020.

althea southeast asia

Google Search Interest – ‘Korean skin care’

Side note on why K-Beauty is popular

To understand why K-Beauty is so popular, it helps to understand the unique dynamics of the Korean market. It comes down to the fact that Korea is a small country with a condensed, demanding and fickle population.

Companies are constantly under pressure to innovate and create new products in order to meet the differing demands of consumers and to stay relevant. In traditional beauty companies, it takes approximately two years to create a new product, whereas Korean companies do so within months. One of the best examples of this hyper innovation is highlighted in the unique ingredients used in K-Beauty products ranging from snail mucus, horse oil to egg whites.

Despite the high demand for K-Beauty, it is extremely difficult for consumers to purchase K-Beauty products in most parts of Southeast Asia. The first challenge is the limited selection – there are over 10,000 K-Beauty brands in Korea but less than 100 of them are available in Southeast Asia.

This is because the demand in Southeast Asia, although growing, is still small relative to Korea’s domestic market and their main export market, China. On top of this, the available brands usually only offers a limited selection of their top sellers, not the whole product line.

The second challenge is pricing. In other parts of the world, K-Beauty is known as a value for money product, but because of distribution agreements and import duties in Southeast Asia, customers often pay up to 100% more for the same product.

For example, Laneige sells its Water Sleeping Mask for 28,000 KRW in Korea but it costs approximately 35,000 KRW if a consumer purchases the same product in Singapore.

Althea-southeast-asia

althea-southeast-asia

At Althea, we saw that the vast price discrepancy represented an untapped opportunity. By hiring a specialized team of K-Beauty experts in Seoul to work with brand owners and distributors to build our assortment and negotiate pricing, we have scaled our assortment from 300 SKUs to 3,000+ SKUs. For brands that we have direct relationships with, they see Althea as an easy way to enter five different markets that have diverse languages, culture and regulations.

Our pricing is comparable to what customers would pay in Korea, and we can generally deliver across the region to most areas within five working days. This has established us as an authority on K-Beauty, which has two main benefits:

1. The majority of Althea traffic and transactions come from unpaid channels (Direct and Organic Search results)

2. It significantly improves cost per order (CPO) –  in our more mature markets, we are getting marketing spend ROIs in excess of 25x

althea-southeast-asia

*Blended CPO over the last 12 months in one of Althea’s more mature markets. Source: Althea cost per order, internal data

2. LOCALIZATION

Even though K-Beauty is considered a universal product, it is still important for us to localize our service and offerings. We localize every consumer facing aspect of our business and operate a different website for each of our specific markets. From changing web addresses to suit local domains such as my.althea.kr for Malaysia and th.althea.kr for Thailand to having website content in the local language.

For our marketing campaigns, we engage 150-200 top beauty influencers in each country to act as local brand ambassadors. We benchmark our pricing by market to ensure we have the cheapest prices in each country we operate in.

Payments has always been a hurdle in ecommerce, we cover most major options ranging from credit card, local bank transfer, over the counter, and cash on delivery, which is crucial to success in Southeast Asia, due to low credit card penetration.

Our logistics operations work with 15 local last mile couriers and five local return centers to facilitate ease of deliveries and returns and we’ve adopted our customer service operations to offer social network support in each market.

Althea’s localization efforts: 

  • 5 websites
  • 4 languages
  • 750+ beauty influencers
  • 70+ payment options
  • 15 local last mile couriers
  • 5 local return centers

3. FINANCIAL DISCIPLINE

We are fortunate enough to be a part of the beauty industry, one of the only industries that has maintained its margins for 100+ years. Our product margins are multiples higher than a typical ecommerce business. Combined with the inherent stickiness of the products and a highly engaged customer base (18-34 year old females), our customer LTV is attractive.

These factors mean that we can afford to be aggressive with our customer acquisition cost (CAC) and other customer centric costs. While we invest aggressively in marketing and customer service, we practice strict financial discipline in all other aspects of our business. This means that while we do business in six countries, we only operate two offices and one central fulfillment center. We have one office in Seoul, which has 10 full time equivalents (FTE), and is responsible for sourcing, logistics, HR, and legal.

Our second office in Kuala Lumpur has 20 FTEs that handle development, marketing, finance, and customer support. This asset light model has allowed us to enter a new market once every three months since launching.

We generate significant savings by not duplicating functions across local offices, and in turn we re-invest those savings into growing our business.

By adhering to these three core principles – focusing on a niche category that we can execute well, localizing all customer facing aspects of our business, being efficient with our capital and a little bit of luck, of course, we were able to enter five markets and build a $10 million a year business within one year. Looking forward, we hope to expand our business to the rest of Southeast Asia as well as the greater APAC region.

althea-southeast-asia

The Althea team at their one year anniversary party

BY DAVE CHANG, HEAD OF MARKETING & EXPANSION AT ALTHEA

It is hardly a secret anymore – ecommerce in Southeast Asia is an enormous $238 billion opportunity that has been under the global radar in the recent years. It’s no longer about whether businesses should have an online presence, but instead how they can stay relevant to their audience in a quickly crowding space.

Indonesia Ecommerce Landscape

ECOMScape: Indonesia details the growing ecommerce ecosystem as of 2016. Source: eIQ

One market that has time and time again stood out is Indonesia thanks to more accessibility to mobile devices, affordable data plans and a youthful demographic propelling social channels and social commerce to the leading activity on the internet. It is a clear goldmine for brands, retailers and investors alike to unlock over 250 million unrealized online shoppers.

So how are they going to do this? Well, Indonesia is a mobile-first country and its citizens update their social networking apps twice more frequently than games and fives times more than music/video apps according to a study by Baidu.

In the next three years, Indonesia is expected to have over 92 million smartphone users, up 67% from 2015. This will push many startups to skip desktop entirely and focus on smartphone friendly ecommerce products. New apps have been the popular way to reach customers, but large spending for development, maintenance, and marketing have restricted companies from finding long-term success. And what happens when app downloads start to slow down as currently happening in the US?

wearesocial-indonesia-eiq

The rising global resistance to new apps

Almost 50% of smartphone users in the US did not install a new app last month while less than 25% of the ones who did returned to it after the first use. What’s even more shocking is a full 94% of revenue in the App Store comes from only 1% of all publishers, think Google and Facebook.  

Mobile isn’t dead but the opportunity is shrinking. So what does this mean for businesses scrambling to capture the attention of the world’s fourth largest population who prefers to use on average only 6.7 apps?

You don’t chase customers, you find them where they already are.

For the first time, messaging apps have surpassed social networks and that’s where chat commerce comes into the story.

Chatbot, chat commerce

Chat commerce isn’t the future, it is the present.

Chat commerce or conversational commerce is the intersection of messaging apps and shopping and is already a very familiar concept in the West. Businesses understand the importance of being readily available to their customers, especially as a poor customer service experience will drive 89% of them to a competitor.

According to Facebook, more than 50 million companies operate on its platform and send more than 1 billion business messages every month.

But having properly trained customer service reps to support hundreds to millions of personal conversations in parallel is difficult to scale for any business. Solution? Chatbot.

A chatbot is an AI (artificial intelligence) feature of a chat or messaging platform that simulates a human conversation with the user in order to provide them with the information or service they’re looking for.

Brands overseas like Taco Bell have partnered with Slack to allow customers to order and pay through the team communications platform.

Think about Siri who has been helping Apple users carry out tasks since 2011 or Amazon Echo, an at-home device by Amazon, which encompasses a chatbot named Alexa to read aloud weather reports, set alarms and more importantly, help customers order new products from Amazon.

chat-bot-amazon

Screenshot from Amazon Echo commercial.

These are only a few of many examples. Facebook Messenger also opened its platform earlier this year for businesses to build chatbots through its Messenger Send/Receive API.

The API will support sending and receiving text and also images and interactive rich bubbles containing multiple calls-to-action.

chat-bot-benefits, southeast asia chatbot

A chatbot can clearly offer a business great benefits to get closer to customers in a medium they are already familiar with, so why has there been little activity in Southeast Asia?

Call all chatbots

Southeast Asia has largely mirrored the West and particularly China in development of its ecommerce maturity. Yet, current chatbot growth has been stuck at the ‘idea phase’ – a lot of chatter and buzz about its revolutionary importance but no product.

“The reason why companies in Southeast Asia haven’t created chatbots isn’t because they don’t think the opportunity is there, but they lack the resources and most fundamentally – AI talent to build it,” comments Lingga Madu, Sale Stock co-founder.

No company has released a true commercial-scale, transaction-enabled MVP, that is, until now.

Sale Stock case study: Facebook Messenger’s first chatbot in Southeast Asia

One lesser talked about company has already begun testing its chatbot with Facebook, Indonesia’s most popular social channel. Sale Stock, the mobile first shopping platform widely popular among Indonesia’s young females, has become SEA’s first company to launch a chatbot that can handle end-to-end transaction on Facebook’s Messenger Platform.

Meet Soraya AI, a relatable, cheery chatbot who handles 100% of queries coming to Sale Stock’s Facebook Page and the brainchild of Facebook, Google, Palantir, and NASA engineers recruited by Sale Stock around the world.
sale-stock-chat-2, southeast asia chatbot

Soraya uses machine learning to shuffle through queries and decide whether to answer it autonomously or give recommendations to an agent instead.

Frequently asked questions such as “do you offer cash on delivery?” or “do you sell high heels?” are replied to almost instantly. In development since 2015, she can already handle 22% of all queries autonomously.

The beauty of machine learning is that the more information she receives, the smarter she becomes and the more accurate her answers will be.

Soraya has already improved response time by 20 – 40 times and currently replies within 60 s. That has granted Sale Stock a response time badge on their official Facebook page.

sale-stock-fb, southeast asia chatbot

Soraya has also been fed large amounts of past Sale Stock customer queries to enhance her intelligence. This combined with recent purchasing behavior and browsing history allow her to recommend consumers personalized products.

Buying the product is even more simple. Soraya asks for confirmation of the item, correct size and color, all within Messenger, and requests address and payment method. If it is a returning shopper, all previous payment information is saved so purchase is simply a click of yes.

sale-stock-chatbot, southeast asia chatbot

“Soraya was created to meet the needs of our customers, many of whom are living outside major cities on limited social media data plans where chat is free but browsing is not. Some have never even been exposed to digital shopping carts but chatting is second nature,” – Jeffrey Yuwono, Sale Stock President.

Trust is also a major concern that holds many Indonesians back from trying ecommerce. By creating a personable chatbot on a familiar channel, brands hope customers will feel comfortable sharing their personal details.

sale-stock-chat, southeast asia chatbot

Chatting with Soraya on Facebook Messenger

Sale Stock chat bot, Indonesia

What’s next for Sale Stock?

The company is already working to create viable chatbots for WhatsApp, LINE, BBM and SMS as they are the most popular messaging platforms their shoppers use. Sale Stock strongly believes in the Lean Startup methodology, “getting it out there as soon as possible to collect real, user feedback”.

“We’re still in the very early stage of our product and ironing out the bugs and adding features iteratively,” comments Madu.

All inquiries going to Sale Stock are monitored, independent of the channel source, on one platform created in house to control flow and fix any arising bugs.

The team hopes to fine-tune its technology to quite possibly launch SaaS in the future.

“The success of chat commerce depends on how well the machine can distinguish the details: context, intention, the slang, mix of dialects, and even the use of emojis so the customer never feels like they are chatting with a bot. The platform has to be robust enough to handle these typos and fringe use cases,” says Madu.

The future of chatbots

There has always been a fear of AI replacing tasks typically performed by humans, but customer support is a tricky area since personalization is at the core.

“As brand loyalty and exceptional customer service become the main priority for brands, companies simply cannot afford for bots to completely handle customer service and risk creating a negative experience. With that said, the live customer service representative will always have a place with the overall customer experience,” says Mayur Anadkat, Vice President of Product Marketing at call center software provider Five9.

The moment has not yet been reached when machine learning enables 100% accurate and instant replies to customers no matter the language, mix of dialect, slang or emojis – but it is in the foreseeable future. AI is here to enhance, not replace.

Not only will the rise of chatbots improve the reputations of brands but it will be expected of businesses by the next generation of shoppers. As Uber product manager Chris Messina put it, bots present a new, unpolluted opportunity to build lasting relationships with people.

Ultimately, the lack of friction is what makes the shopping experience a pleasant one and what will drive the A players to the head of the game.

By: Cynthia Luo

When renovating spaces in Jakarta, always be aware of the flood season risks!

Despite the onset of banality of the term ‘world class’ (which, in Southeast Asia, seems to signify that something functions adequately without posing gross health and financial risks) my team and I sought to reinvigorate the term when setting up an ecommerce warehouse in Indonesia. What ‘world class’ meant for us was to design an ecommerce warehouse that could sustainably handle a 200% uptake in order growth. Our hot startup ecommerce client has been enjoying an intense surge of ecommerce success and needed long-term scalability, fast and not rely on sloppy patch-work processes typical of fast-growing startups everywhere.

When setting up an ecommerce warehouse in Indonesia, the targets were the following:

  1. Exceed our competitors in terms of magnitude (400+ employees)
  2. Scalability (operational capacity triple initial capacity)
  3. Innovation (semi-automated conveyor system and technology)
  4.  Customer experience (detailed project updates and account management).

With my seven years of solution design and project management experience with notable German global logistics providers, my obsession with efficiency and quality (not only a German stereotype) and a scrappy and clever team, I didn’t think it was out of our reach. But by no means would it be easy.

Our learnings for industrial engineers & logistics PMs in emerging market ecommerce

My advice to young solution designers, industrial engineers or logistics project managers in the ecommerce market, the devil is in the details –critical information gathering is what differentiates one solution designer from the other. A successful implementation depends primarily on the level of details the solution entails.

Here is a recap of our trials and tribulations to setting up said world class fulfillment center in Indonesia.  While not totally comprehensive, you will learn about the key considerations and parameters in crafting or implementing your next ecommerce fulfillment solution or project.

So what was in store for us? First we had to build a complete new warehouse, redesign all existing processes and hire manpower, all within a six months’ timeframe. The challenge was not just in pure size of the project but the tight deadline we were working within.

1.Sizing up the problem – order volume projections

The first thing we did was get our Solutions Design Team assembled to start analyzing historical data against projected volumes. Using detailed solution templates and tools, we submitted a proposal simulating the required warehouse space, storage solution, resource requirement and process maps for the operation.

The expected growth calls for a warehouse design that meets both storage-picking optimization and long term scalability requirements. The projected sharp increase in handling volumes also meant that we had to reassess the current technology and operational processes that could no longer support the projected growth in transactional volume.

2. Choosing the right warehouse & location

There weren’t many options from a real estate perspective as we needed to find a ready space of approximately 7,000 sqm in the heart of Jakarta within two weeks. Fortunately for us, there was an old carpet manufacturing factory in Cawang (East Jakarta) that was winding up and looking to rent out the space immediately.

It was the perfect site based on its location, size, availability and price; it even had a mezzanine floor which could be transformed into offices for our client.

setting up an ecommerce warehouse in Indonesia

Interior of the factory.

3. Redesigning the space to be ecommerce optimized

However, it being a factory, the specifications of the building structure were not meeting our initial warehouse design. Some key areas that we needed to fix asap included the low height of the factory, damaged flooring, excessive interior configurations, poor network and electrical systems etc. The factory was also positioned on a lower ground right next to a river bend making it highly vulnerable for floods.

Often mistaken, the key criteria in designing an optimized storage solution lies in the pick profile of the operation instead of its inventory profile.  

The solutions team started plotting the space requirements based on the blueprint of the factory. The proposed design included a 2500sqm 4 level mezzanine storage structure equipped with 2 cargo lifts, a semi-automated conveyor system, an office space to accommodate up to 150 office staff as well as a state-of-the-art studio room for photoshoot.

The proposal also highlighted the required renovation to the original state of the factory such as roof elevation and demolition of existing columns and rooms.

setting up an ecommerce warehouse in Indonesia

Before beginning, we developed layout design and work flow simulation.

The main focus for the warehouse setup was the roof elevation since nothing can be done until the roof is up. Unfortunately, a considerable amount of waiting time was invested to governmental approval and regulatory certification. The actual roof reconstruction works were then impeded by rain- each time it rained the project paused and when the rain stopped, had to wait for drainage drills to kick in before continuing the construction.

4. Selecting the contractors & sticking to timelines

With the approval from the customer, we proceeded to seek quotations from warehouse vendors. The vendor management process included our assessment and recommendation based on the solution designed as well as the customer’s budget. After almost a month of negotiations, three vendors were selected to participate in 8 key areas of this mega build out.

setting up an ecommerce warehouse in Indonesia

Warehouse setup deliverables

With the sign off on the overall solution, proposal and investment budget, the project implementation timeline started. All in all, we had 3 months to transform the old factory into a “world class” facility. As most activities were interdependent; for example, activities within the factory needed to hold up until the roof elevation was completed, the mezzanine buildout could only start after the required floor repair and site demolition were completed – timeline management was particularly challenging.  

Once the project timeline was set, weekly meetings were held with all vendors as well as key project stakeholders to make sure the timelines were met and that all potential risks raised were mitigated.  

Fortunately, it didn’t rain much during the construction phase and the skeleton of the warehouse was completed on schedule.

setting up an ecommerce warehouse in Indonesia

Before and after the roof construction. This was one of the toughest parts due to weather.

5. Building the inside

The setup proceeded with parallel work on mezzanine installation, workstation setup, office buildout, roadworks and electrical setup. It was a good learning experience for the team as we participated in electrical distribution layout design, network cabling blueprints as well as water reservation and supply including the number of toilets required.

Just when we thought the rain could no longer serve as a threat to the project, a heavy downpour in the middle of the night coupled with a broken dam leading to the river resulted in a flood in the warehouse through the overflowing reservoir at the back of the warehouse.

setting up an ecommerce warehouse in Indonesia - jakarta floods

When renovating spaces in Jakarta, always be aware of the flood season risks!

6. How we dealt with the Jakarta floods

This led to an additional project milestone on enhancing flood preventive measures. Corrective action plans were immediately put in place. A response team was formed to drain the water out of the warehouse – sandbags and temporary floodgates were installed in preparation for the next heavy down pour while deliverable items such as enlarging reservoirs, renovating the walls as well as installing additional drainage facilities were carried out in phases.

Fortunately, no assets were damaged during the process but the overall timeline definitely took a hit; since then, a rain-phobia was developed throughout the project.

Other challenges such as the overall safety of the mezzanine as well as the levelness of the conveyor led to several assessments and trial and error to achieve the most optimal setup for the operations. By mid May, the warehouse was completed.

setting up an ecommerce warehouse in Indonesia

Cawang fulfillment center completed.

7. Operating the warehouse – hiring the manpower

Using productivity numbers collected from existing process cycle timings and calculated assumptions made on new operation processes, we simulated how many people we would need to handle the projected volumes.

The resource modelling tool allowed us to define clearly the number of manpower required by positions, functional roles as well as shifts – making it easy for hiring. The challenge however was to have all 408 headcounts qualified, interviewed and hired on time for training, migration as well as Go live.

We overcame this by engaging with local established headhunting firms and manpower agencies and successfully hired 80% of the required manpower with the outstanding positions filled out by the existing team to add experience and stability to the run operation. Unfortunately, the hiring process was not all straightforward, replacement or rehiring processes was an ongoing procedure due to natural attrition and absentees.

It is important to note that an influential warehouse manager is the key to staff retention. An effective warehouse manager creates strong family-like bonds within the team, develops a team of trusted leaders and is respected by all.

Two key responsibilities of a warehouse manager are 1) Ensuring the team’s commitment towards go live and 2) Sustain operation excellence after Go live.

8. Training & knowledge transfer to local management

There will come a stage in the project where the project manager will start transferring ownership and controls to the warehouse manager. The warehouse manager will take over on training deliverables, shift planning and overall discipline of the team.

The team sat down to plan out the rotational training schedules and deliverables targeting to get all 408 employees ready for the launch- Go live. The topics covered product knowledge, system processes as well as hands on operations in the existing and new warehouse. Each employee was assessed on what the topics covered at the end of each day to ensure that we have everyone in sync before we moved forward.

Workstations readiness and network connectivity were stress tested during the User Acceptance Test (UAT) exercise to ensure all potential risk were managed prior to go live. To curb with the intensive operational requirement, we had installed a backup line for network as well as power distribution.

The last week of training was held in the new warehouse where dummy orders were created to simulate end to end operational process in the new warehouse. There, situational tests were created to assess the employees’ response and rectification initiative.

9. Migration challenges & Jakarta’s travel ban

The lead-time given for migration from old to new location was five days.  Using the same solution templates and tools, we were able to define the targeted letdown and the number of trips required per day. These targets were translated to the team to ensure we stay on track for each day.

The main challenge during migration was the travel ban that only allows vehicle movements after 9pm. As such, coordination between the operations and transport team were crucial to ensure we optimize the letdown and loading process effectively before the travel ban is lifted.  

Other considerations such as inventory accuracy, quality assurance and system management were key to the successful migration. All in all, approximately 1.4 million units over 90 trips were transferred.   

Setting up an ecommerce warehouse in Indonesia

Migration process of 1.4 million items to the completed center

While there is no perfect project where all factors can be foreseen when setting up an ecommerce warehouse in Indonesia; experience and dedication are what minimizes that inadequacy and in order to get there, it takes no less than pure focus, passion, structure, top management commitment and hard work and maybe some luck with mother nature.

By Mitch Bittermann, Group CLO and Kenneth Thean, Regional Director of Solution Design at aCommerce