Unique Research · Interview · April 2, 2024
Background
On January 24, 2024, Unique Capital and partners held the 2024 AIGC Application Development Summit in Shanghai, with nearly 1,000 AI practitioners attending. This piece is edited from the guest keynote / roundtable segment.
Guests
- Guo Xiankai (郭贤凯) — Founder, SaleSmartly & HelpLook
- Huang Zhen (黄臻) — Founder & CEO, AroundDeal
- Chen Can (陈灿) — Co-founder & CTO, FOSHO
- Xu Xiaoxi (徐晓熙) — VP, Yinxiang Tech / Impression Tech (印象科技)
- Moderator: Zhu He (朱鹤) — Principal, GOO
Trends roundtable
How each company stepped into the “go-global pit”
Q: Please introduce yourselves, then focus on how your companies stepped into the “go-global pit.”
Xu Xiaoxi, VP, Yinxiang Tech: One of Yinxiang’s core products is Yinxiang Biji (印象笔记) — a brand most people know well. For us, entering overseas markets is not simply “going out”; it is a normal step in a globalization strategy. Our background is unusual: we originally came from Silicon Valley’s Evernote brand. So we inherited Silicon Valley leadership in brand, user base, reputation, and tech foundation — a path from Silicon Valley to China, then from China to the world.
Globalization DNA was rooted in our culture from day one. Many team members are from Silicon Valley or educated in Europe/US, so we stay closer to the global ecosystem — that was why we chose this path.
After years of product iteration we have built a soft+hard product line including an AI-powered matrix. In project kickoff and R&D we always balance domestic user habits and internet ecology, but on tech and product design we stick to global standards — so we accumulated products that can face global markets directly, and now keep innovating to push them out.
Chen Can, Co-founder & CTO, FOSHO: When we chose the founding direction, overseas expansion was first choice — based on the team’s prior overseas experience. So from day one we decided on a go-global strategy.
I also flip the question: at founding, did we ever consider not going global? We saw many Chinese firms choose overseas for growth. We realized localization, intelligence, and related factors would shape Chinese brands’ future paths and ecosystems.
AI accelerated that process and opened infinite opportunity. In China, “affiliate marketing” is not a very mature field — a huge opportunity and runway for Chinese brands. We seized it and launched matching products and services.
We were glad that in 2023 more customers and investors chose us with confidence. Many peers also picked this direction. Go-global may not be the only choice, but it is at least one widely recognized path.
Huang Zhen, Founder & CEO, AroundDeal: AroundDeal focuses on global business data, helping B2B firms grow globally — roughly an overseas version of Qichacha, centered on data and marketing. That product+data combo makes us naturally push globalization — our first advantage.
Second: my prior venture was in exhibitions; from 2015 I explored globalization and built years of B2B global marketing experience. So in the second founding I again chose globalization as the direction.
Guo Xiankai, Founder, SaleSmartly & HelpLook: I have always dreamed of a SaaS factory shipping multiple SaaS products. Since founding in 2019, after mastering some SaaS playbooks and marketing tricks, I decided to apply them to more products for validation. At first we mainly served domestic users who used overseas SaaS but were price-constrained. Customers asked for that; we shipped similar SaaS at lower prices — e.g. products that cost thousands of dollars sold at $100.
At first we did not obsess over pricing; we prioritized meeting needs. Building, we referenced overseas competitors and found multilingual support was not hard. Adding language packs, we supported Chinese, English, and Russian at controllable cost. Then we put multilingual on the site and product UI. At that time we still only served Chinese customers — mainly cross-border e-commerce and traffic go-global. Some paid via PayPal or bank cards, so we integrated both.
After half a year of hard improvement, more customers recognized us. SMB customers recommended each other — even Chinese customers talking with overseas, Russian, or Southeast Asian customers.
I personally did CS and product for two years. You need to know what customers need before you know what to build. Especially in B2B, knowing your customer is hard; what you want to build may not be what they need. We put 70% of product R&D into meeting customer needs and strictly forbid subjective demand invention. Many overseas customers gradually chose us spontaneously; overseas users now exceed 50%. So we did not actively “go global” — as the industry and product scenes developed, we attracted global attention.
We shipped Ads Power for multi-account marketing. Marketers doing content or ads often need many accounts at once; Ads Power browser makes multi-account login easy for marketing work.
During build, customers suggested unifying customer communication in one UI instead of switching browsers. So we shipped SaleSmartly — sales and CS can talk to all customers on Facebook Message, Instagram, email, WhatsApp, and more in one place, sharply raising communication efficiency.
Yinxiang returning to the US vs Evernote
Q: For Yinxiang’s products counter-attacking back into the US market, what differentiation vs Evernote itself? What experience from China? How does the US differ from China?
Xu Xiaoxi, VP, Yinxiang Tech: Our product line covers not only a software matrix but also smart hardware and cultural peripherals. On AI we built our own advanced foundation model and, via routing plus multi-model hybrid drive, power both the software matrix and smart hardware.
Vs Evernote: Evernote was only Yinxiang Biji’s original single software product. Over time Yinxiang’s features and development have far surpassed Evernote. When expanding internationally we first reviewed the full considerations for designing AI products and locked six key elements — the model as core power matters, but so do deep understanding of user need scenes, choosing the right interaction and carrier, and how to use different data. On data we especially protect privacy: which data may be used for training/inference, which must be strictly protected. After synthesizing those, we clearly defined the product direction.
On AI enablement, software and hardware have different emphases. Especially in smart hardware, the goal is a “second brain” that helps users remember, manage knowledge, and turn knowledge into action. In practice a single software may not cover all needs, so we shipped AI-native smart hardware adapted to the stack: EverMARKER book-excerpt pen, EverPEN stylus, EverPAPER e-notebook, recording earbuds, and more.
Those hardware products are designed from user need and scene; interaction and carriers were carefully rethought. Overseas, user habits and internet ecology differ from China, so we keep adjusting. E.g. a light-recording product we shipped domestically is mainly used as a personal tool like Weibo or Moments. Overseas — especially early SaaS users like KOLs — care more about sharing to TikTok, X, Facebook, Instagram; we are integrating with social apps to strengthen sharing.
FOSHO overseas vs domestic — special adaptations
Q: For FOSHO’s overseas version, how do overseas customers differ from the domestic version, and what special adaptations did you make?
Chen Can, Co-founder & CTO, FOSHO: Domestically, affiliate marketing is not mainstream and is a small share. Overseas it has 20+ years of foundation and is important for many companies — the “mid/small channel” that can contribute up to ~30% of revenue.
Given industry traits we usually cannot connect to customers directly — so we need not only customer feedback but also proactive market entry to discover more problems and needs and redefine the product. Our redefine method: two words — 打透 (drill through) and innovation.
Overseas we often used dispersed guerrilla tactics that never fused with mainstream marketing — so no complete marketing ecosystem. Entering this field, we successfully combined our resources with the whole marketing ecosystem: big-media high exposure, blog recommendations, content marketing, price comparison, plus small resource sites. That ecology is common in China; overseas we were among the first to truly achieve it.
Though still a small company, with knowledge/experience plus AI help, we used few resources to do what had not been drilled through before — a big step. After two years of self-R&D we shipped a new advanced affiliate-marketing AI cloud covering insight, decision, transaction, management, and optimization across the full chain.
Innovation mainly comes from needs unmet by existing products. A shared industry problem: products are too complex and communication extremely cumbersome. We shipped an intelligent-interaction product so beginners can master needed functions. E.g. when a user asks where a feature is, the interaction guides clearly; when a need is unclear, it dialogues to clarify and offer solutions. That intelligent interaction is the core of our AI innovation “3+1 strategy” and has shipped — solving ToB customer-interaction pain and raising satisfaction and efficiency.
We also shipped FOSHO AI Spark for small affiliate customers with unclear needs — generating materials from their copy and visual style in one click. What used to need long workshops or brainstorms is now one click — accelerating from negotiation to publish.
We also proposed “super alignment” to address instability and reliability of AI output in commercial apps — reflecting on “super misalignment.” Our team took a different path from Ilya’s. Super alignment is extremely hard; super misalignment focuses on raising commercial quality. We think that is an innovation worth debating.
AroundDeal’s go-global path
Q: AroundDeal’s product looked international from day one — can you talk more concretely about your go-global path?
Huang Zhen, Founder & CEO, AroundDeal: Domestic vs overseas business data differ at the root. Domestic business data mainly sits on industrial-commercial registration data; firms extend different businesses from that base. Overseas logic is not built on that registration base — first difference.
Second: overseas marketing logic differs from China and focuses more on small-B or large-B markets. E.g. in domestic ToB, people may rely more on AI Call for quality reach — even at 1‰ conversion they go all-in. Overseas that often does not work. Overseas AI Call is more for CS and more toward the back end. From those two angles, domestic and overseas business-data DNA are completely different.
Before AroundDeal commercialized we went through adjustments. First we built a product designed for China, then abandoned that v1 and built a wholly new AroundDeal. Internally we redefined the product for a global standardized experience not limited by language. We borrowed many US peers’ products, absorbed their standardization logic, and finally commercialized successfully.
Challenges along the way: overseas customers usually have strong self-serve ability; domestic users need more external support to use the product efficiently. We optimized the product to better fit domestic habits and simpler ops — a key problem that must be solved when bridging the markets.
Marketing & GTM that fit overseas
Q: What overseas-fit marketing methods and GTM means has everyone tried?
Huang Zhen, Founder & CEO, AroundDeal: We have tried many means: conventional SEO/SEM — outsourced or in-house; LinkedIn and Facebook ads; and the affiliate programs mentioned above. Since H2 2023 we have nearly 70 affiliate partners globally — a welcome result after half a year of exploration.
On ads, because we own data assets, we maximize that bottom layer. E.g. on LinkedIn we do not use LinkedIn’s audience packs directly; we use our own data for precise targeting. Entering Singapore, we filter Singapore B2B sales managers from our DB and upload to LinkedIn — raising ad ROI.
Our methodology: combine bottom-layer data with LinkedIn ads. Ad volume may not be huge because ads need exposure and LinkedIn match rates are limited — e.g. even with 100k/200k/300k people in an audience, only about 70% may match and get exposure on LinkedIn.
Most effective for us: use our own bottom data for networked marketing reach. As a seller we filter Singapore sales managers from our data, import them into our “customer engagement” module, and let the system automate order pushes.
Chen Can, Co-founder & CTO, FOSHO: Our team has 10+ years overseas, mainly C-end. I also had B-end experience elsewhere, so I know common consumer channels. But once truly in B2B, practice shows differences. Beyond traditional ToC big media, big exposure, PR, and social, B2B needs different strategies.
For startup brands, SEO/SEM may show little effect at first, but starting early is wise. They won’t pay off instantly, but over time and experience they become among the most effective tools.
Choosing the most effective marketing should be based on company stage and market environment — no one method fits all. Given current competitive pressure, CPS (pay-for-performance) is worth trying — earn first, pay later, raising ROI. Even experienced firms can err; we must watch and avoid common traps.
In the past we used the same strategy and product-push methods at home and abroad. As you said, we once thought multilingual was just simple translation and pushed the product overseas. Practice proved overseas customers’ needs, service standards, and habits differ sharply from China — simple copy-paste does not deliver expected results.
Xu Xiaoxi, VP, Yinxiang Tech: Our product is distinctive, and we firmly believe only product-led growth (PLG) can efficiently reach our C-end consumers. Smart-hardware involvement also gave us offline-channel insight.
First: many domestic founders/entrepreneurs succeed more often in smart-hardware go-global — they fully use China’s rich supply chain, so hardware can validate at near-software speed. Conversely, in software — especially SaaS in an adjustment cycle — domestic founders may not be close enough to users, or know overseas culture and use scenes well enough.
Though our brand has years of domestic ops and 80% of user growth is organic, internationally we are still a new brand; most customers come via PLG. We partner with platforms including Product Hunt, Google Accelerator, and long-term good cooperation with Apple and others.
On hardware, crowdfunding via Indiegogo and Kickstarter has also delivered solid results. In current marketing practice we think GTM must care not only about strategy and channel choice but also early product positioning.
On smart-hardware GTM we use similar methods. Facing strong competitors we consider a value strategy — lower price, higher value. For some commodities-style products we choose pure volume, not obsessing over retention — convert and charge directly; simple and direct.
Guo Xiankai, Founder, SaleSmartly & HelpLook: Two very effective marketing modes. First: content marketing — on-site and off-site. On-site needs quality content, keyword optimization, attract users and convert. Off-site mainly uses backlinks to raise site authority, exposure, and impressions.
Second: partner marketing — very similar to FOSHO’s product. Ours has three angles. First, cross-industry partners: each year find 100 partners with similar audiences in each country, then mutual friend-links and blog posts to raise exposure and awareness. Even if customers don’t feel it directly, the cooperation has value for them, so we always do it.
Another key: seek many influencers. Our product fits small-B marketing or ads scenes, and the small-B market has many influencers — we gift product, have them make promo videos, and put affiliate links under the video. Influencers get free product plus affiliate rewards.
Our biggest early challenge: when first going global we did not segment regions and wrongly thought we could run globally in sync. That strategy grew nowhere. Later we split trials, focused on the few countries we thought most promising — e.g. Russia and Vietnam — and hired locals for marketing, BD, and local service. After the shift, performance grew significantly — monthly growth of 10% or 20%.
— END —