3 views: Predicting 2023’s key startup themes

A good way to be wrong is to predict the future. A good way to be incredibly and embarrassingly wrong is to predict the future in a medium that is public and everlasting. With that in mind, welcome back to another episode of “TechCrunch Predicts.”

As we did last year, TechCrunch’s Natasha Mascarenhas, Anna Heim and Alex Wilhelm are back with a passel of predictions, hoping to percolate new postulations in your ponderer. Not all the below will come true, but it should help explain where our heads are at after a year’s reporting, writing, newslettering and podcasting. Among the three of us, we’ve spoken to hundreds of people this year, giving us — we hope — a modicum of insight into the state of technology today and what could be coming next.

When we went back and looked at our 2022 predictions, we got things more right than we could have expected. So let’s do it again! What do we see for 2023? Anna is betting on API-led startups, Natasha is wagering on silence and Alex tripled down with a trio of predictions that, we presume, will be wrong in short order. Enjoy!

Anna Heim: The rise of API-first startups will continue in 2023

I am convinced that API-first will be a major trend in 2023, with this approach being both more widespread than it was previously, as well as more successful than less API-heavy options.

That APIs are on the rise isn’t exactly new — but API-first startups are a subgroup in this world, and one that is enjoying tailwinds.

According to findings from API platform Postman based on internal data and a survey, companies that report spending 76% of their development efforts on APIs are “35% more likely to forecast an increase in hiring over the coming year than respondents ranking themselves lowest in API-first.”

Coincidentally, the share of companies that embrace an API-first approach is larger than a couple of years ago. Just like in 2021, Postman found out that two-thirds of the respondents “ranked themselves as a five or higher in terms of embracing API-first” — compared to 62% in 2020.

Venture capital firm GGV Capital is also bullish on this trend, and it built an index that gives us further data points on how things are going. During the third quarter of 2022, GGV’s API-First Index “increased from 50 companies to 60+ companies that have now raised over $50 million,” the firm said.

It is worth mentioning the 12 companies that GGV last added to its index: AssemblyAI, Belvo, Deepgram, Fabric, Fauna, League, Lokalise, Patch, Sardine, Solid, TrueLayer and Truework. The list itself is an interesting data point because it goes to show the diverse range of companies that are adopting an API-first philosophy.

While APIs are obviously well suited for fintech, especially in connection with open banking, we have already witnessed how they expanded to adjacent fields, such as HR and insurtech (through embedded insurance.)

In 2023, I expect API-first companies to become a fixture in health tech and also connect the dots with two other major trends: headless commerce and low code.

It is often when trends collide that they become more powerful, so I will definitely be looking out for these combos next year — and will also have an eye out for GGV’s Q4 updates to its index.

Natasha Mascarenhas: A different type of vulnerability — silence

News is built around tension. Sometimes that could be in the form of success despite all odds, and other times it could be about one of the world’s largest crypto exchanges falling apart in a matter of days. Regardless of whether the tension highlights a character in a positive or negative light, it likely tells us something sharp about ourselves and the industry we operate in.

Which is my long way of saying that 2023 may bring a different type of vulnerability to entrepreneurs. The past 12 months have popped tech’s bubble in a way that has likely inspired an entire generation of founders to build differently and more thoughtfully — if not for investor money and press badgering, but for the long-term loyalty of customers. To me, 2023 will mean the death of the splashy valuation headline and the rise of revenue-oriented metrics.

I realize I’m kind of manifesting my ideal outcome here, but it’s not all self-serving. I don’t think seed-stage companies are going to tell their valuations to TechCrunch. Instead, I think we’re going to see the rise of stealthy companies that take their time in being ready to tell their story. The media may hate reporting on shadows but admire the idea of entrepreneurs holding their cards close until they have an actual story to tell — not just that they convinced a group of capitalists to give them double the valuation.

The vulnerability from founders we saw in 2022 was different from the vulnerability we saw from founders in 2021; this year was all about apologies for overhiring and a focus on focus, while the year prior was about screaming through pandemic pivot plans and the ability to raise money. Next year may be somewhere in the middle: Founders aren’t going to be showing off their unicorn status because they are more aware than ever before of the stresses that come with hitting the growth stage.

Instead, their vulnerability may take the form of silence. So when they are ready to chat, they may be excited to tell us about ARR, the mistake that unlocked a new path forward, or, hey, that pivots aren’t really that taboo after all.

Alex Wilhelm: Generative AI, SaaS trends and what’s ahead for NFTs

Last year I made a somewhat boring prediction that open source startups would have a big 2022. Wow Alex, so brave. Very daring. Much bold. I promised that I would be spicier this year when we graded our trailing predictions post.

In keeping with projecting a future that could quickly be proven wrong, I have done us all the service by including three predictions so that I can be wrong thrice in the coming spin ‘round the sun:

  • Generative AI in its current and next iterations remains mostly a toy and not yet a product. Look, I like having fun as much as the next person, but mostly what we are seeing with the current generation of generative AI is a series of increasingly impressive toys that do not yet rise to the quality level required for mass commercialization. There is probably a revolution coming; it probably has lots to do with computers doing much more creation for us than they do today, perhaps especially so in the realms of writing, drawing and programming. But not yet. Smart toys are a great stepping stone to the future, but until they can do much more — and more precisely — they will remain more recreation than revolution.
  • The theory that companies are infinite software sponges is wrong, and it’s going to get harder to sell software in 2023. The old maxim of “find a place where someone is using a spreadsheet and build a software tool to save them time” as a way to figure out what product to build a startup around is dying. Why? Because we are suffering from mass tooling overload. Log into your company’s Okta page or similar and look at the sheer number of apps that are present. How many apply to you? Very, very few, I would guess. I am not sure if we are going to see more companies decline new software services or constrain their seat count to prevent bloat — and cost! — but I think that the SaaS boom taught us an incorrect lesson. Yes, companies will buy lots of software. But over the past decade, it seemed like a quantity with no ceiling. It was hard to know that the original conjecture was wrong, as money was cheap during the SaaS golden age and companies were hungry for growth. Now, with cost control being the leading sage of the day, things are going to change, and it will truly separate the startups that are selling something critical from the ones selling something simply neat.
  • NFTs remain in winter until either crypto returns to prior heights or someone figures out an actual use case. At some point, we’re just being mean to the NFT market, but it’s worth sticking to the topic so that way we don’t lose track of it. The NFT hype cycle came and went. Much like the SaaS boom, it was predicated on a period of low capital costs but mixed with COVID-era levels of onlineness that were never going to last. For the NFT asset category to rise again, I think something dramatic will have to happen. A massive, rapid, broad appreciation in the value of crypto tokens, for example. No one likes to gamble more than someone walking by the craps table with a pocket full of house money. Or, you know, NFTs finally, actually, at long last, become useful. But I think the former is more likely than the latter, and that both won’t happen next year.

I look forward to eating my words by March!