Editor’s Notes: In this post, the CEO talks about Where Invisible has come from, and where it’s going.
Wednesday, 05 December 2018
The Strategy For 2019, 2020 and Beyond.
Where have we come from, and where are we going?
First, where we’ve come from:
When we started the company in October 2015, it was a $10,000 a month virtual Chief of Staff service for CEOs. We thought nobody would sign up, but we soon had 5 clients and a waitlist. We raised $500K to execute. But we had no idea how to execute work efficiently, so we couldn’t keep up with demand from our existing clients, or add new clients. That model failed within six months. In Chapter 1, we discovered demand.
Then, for over a year, from April 2016 to August 2017, we worked with just one client. We built over 50 processes to structure our work and make it efficient. We built a labor network to do the work cheaply. And we built Version 1 of our Digital Assembly Line to connect all the pieces: the client with the bot, the bot with the agents, the agents with the processes. In Chapter 2, we discovered supply.
Then in September 2017, we re-launched with new, hourly pricing as a process assistant, not just for CEOs, but for anyone in the organization. We quickly grew to hundreds of clients, delegating thousands of hours a month of work. To keep up with demand, we had to hire and train more agents, and continually upgrade our systems. We built a partnership model, hired more partners to build our systems, and raised $2.1M more. In Chapter 3, we grew.
In July 2018, we began to prepare the model for scale. We created marketing, sales and customer success teams, and began to figure out our cost of sales and support, and the lifetime value of our clients. We added pay-as-you-go pricing to lower the barrier to entry for new clients, and we added subscription pricing to stabilize our revenue. We improved our margins by raising starting prices and increasing our operational efficiency. We automated our least efficient internal processes, like agent pay. We started to build scaleable hiring and training systems. We improved metric gathering and reporting, built a financial model and proper financial systems, and created dashboards to track project management across all teams. We designed, built, tested and iterated on Digital Assembly Line 2.0, and prepared to migrate. We designed and built a Client Web and Mobile App, and prepared to launch it. In Chapter 4, we prepared for scale.
Now, where are we going?
In January 2019, we’ll be entering a new chapter. The theme of this chapter will be accelerating growth.
Now that we’ve got healthy margins, there are two remaining blockers for scale: figuring out our growth formula, and making sure our operations can meet increased demand.
On the demand side…
By the end of December, we need to have clarity on our CAC and LTV, and as long as the ratio is at least 1:2, we should start steadily increasing spend, while running improvement thrusts to lower CAC and increase LTV. By the end of Q1, our spend should have multiplied and our ratio should be at least 1:4. The key is showing acceleration. Investors invest in acceleration. Our rate of growth should be double-digit month-over-month, which compounds and is investable, and ideally, there would be growth in the rate of growth, which is very investable. The most likely channel for this is going to be performance marketing: Google and Facebook ads. As the number of daily MQLs increases, we need to improve our website to convert a larger percentage of traffic — we’re targeting 20%. And the more leads we generate, the more strategists and relationship managers we need to hire to ensure customer success from onboarding through to increased engagement, referrals, testimonials and upsells. Again, this entire machine needs to be well oiled by the end of Q1, sustaining accelerating growth.
In Q2, we’ll hire partners to run brand marketing, content marketing, social media marketing, PR and enterprise sales. And hopefully more talent to take UX and performance marketing — including website, packaging and pricing optimization — to the next level. So by the end of Q2, all aspects of sales, marketing and customer success should be in full swing, and by the end of Q3, we should be hitting yet another growth inflection point.
In summary, the basics of demand generation need to be working by the end of Q1. Then we can substantially increase our investment in Q2, and go from a single-engine machine to multi-engine machine. In Q3, we should be flying on all engines. I can’t overstate the importance of making this happen. It will drive both fundraising, partner pay, and hiring.
We’re counting on accelerating growth to allow us to become profitable and increase our partner pay to sustainable levels by Q3 2019. Otherwise we are at the whim of investors, and capital markets may be in recession. If capital markets go into a recession, the business is very well positioned to outperform — a recession would probably be incredibly bullish for Invisible, because it would drive cost-cutting and efficiency investments, and for both we would be an ideal partner for any business. But if we’re not profitable, we’ll need investment, and in a recession, capital markets dry up.
On the supply side…
By the end of January, migration to Digital Assembly Line 2.0 will be complete, and by the end of Q1, we need to demonstrate that it has made our operations significantly more efficient, and given us a foundation for scale. By the end of Q2, we should be flying. The measure for this will be process efficiency: our speed benchmarks should improve across all processes.
Also by the end of Q1, the Client App experience needs to handle all of the basics well. By the end of Q2, it needs to give clients dynamic control over their processes and preferences, and make it easy to delegate more work in a structured way. The measure for this will be the efficiency of the customer success team: they should be spending more time on advanced problems — building relationships, driving usage, handling complex delegations — and less time on basic problems, like answering questions that a dashboard could answer.
Process Automation needs to begin in Q1. By the end of January, our entire process library should be on the website with public speed benchmarks. By the end of Q1, our internal speed benchmarks should have significantly improved relative to January. By the end of Q2, the should have significantly improved related to the end of Q1, etc. On average, I would like to see at least a 2% per month across-the-board process efficiency improvement, improving process efficiency by 24% per year. This should drive our margins and cost competitiveness.
Then there are the ten thousand little things that go into preparing for scale. Preparing for scale began in Chapter 4, but it needs to continue in Chapter 5. For example, we need to make upgrades to our security, to our dashboards, to our training systems, etcetera x 10,000.
At this point, our partnership should still be small, not larger than 40, but our number of agents is probably between 200 and 400.
By Q3 2019, we should be entering a new chapter. The themes of this chapter should be beginning to scale, improving the foundations, extending the moat.
If we’ve successfully demonstrated accelerating growth in Chapter 5, then we’ll raise up to $10M in a Series A by Q4 2019. And by this point, all of our core systems should be at Version 2 or higher. Our Growth Machine should be at Version 2: that is, sales, marketing, and customer success should all have strong leaders and strong machines that are upgrading themselves every quarter, resulting in accelerating growth. Our Finance and HR Machines should be in a similar position: which means that every quarter our hiring, training, evaluations, performance measurement, finance and analytics systems should be improving. Our Operations Machine should be similarly improving. Digital Assembly Line should be at Version 2.5 or higher, and our Client Web and Mobile App should be at Version 1.5 or higher; both of them should have their next major versions designed, with engineering underway, preparing to launch in Q1 2020. Process Automation should have significantly improved our margins in 2019, and the outlook for 2020 should be rosy: over time, our margins should approach software margins, and nobody should be able to compete with our price per result.
So what do we do with the money? First, we make sure that the partnership is strong. We make sure that we’ve got the right incentives in place. And we identify missing skills and experience, then hire for it. This probably means making more senior hires, now that we’ve got the cash to attract people for whom equity alone was not sufficient. It will be important to have a compensation framework in place that does this successfully while rewarding the people who took early risk.
Second, we increase spend on sales and marketing, and continue to work on increasing our CAC to LTV ratio, ideally to 1:10. By this point, the absolute numbers should be getting quite large. A senior marketing executive needs to be in place at this point.
Our sales team will need to grow significantly, and include more senior sales people capable of doing enterprise sales and deeper expansion efforts. This is going to require a senior sales executive.
The top-line sales goal in this chapter is to grow from a run-rate of between $1 to 3M per year pre-Series A, to about $10M-30M per year through a combination of account expansion and customer success, large enterprise sales, performance marketing and other marketing efforts.
Hiring and training needs to keep up with all of this. In this chapter, the partnership will need to grow from 40 to about 75, and the number of agents will need to grow from about 300 to 1000. This is going to put significant stress on tech to make sure we can scale without increasing our management overhead. And on HR and Finance to make sure that our culture, evaluation, incentives and payment systems scale.
The Finance team will need to make sure that as we grow the business, our overhead isn’t eroding our margins, but rather, that our margins are on trajectory for sustained increase.
For example, at this point, it makes sense to spend money on more off-sites, to bring a remote team together. It will be challenging to maintain Sun Tzu-style asymmetric formlessness with a completely distributed team. If HR is doing its job right, then clusters and hubs will have formed in most major cities, so partners and agents will be able to co-work at will. But paying for flights and accommodations to get the right groups of people together on a quarterly and yearly basis will be expensive. This is the kind of overhead expense that can erode margins, so it should be accounted for. If it is a structural operating expense, it shouldn’t just be seen as part of the spoils of a $10M Series A. If the Series A is any smaller, even more caution and conservatism will be required. The most sensible way to allocate spend to this is to use our profits to build up a warchest, create sub-funds in that warchest for expenditures like this, and create rules around withdrawals from these sub-funds, so that they grow over time and are never overdrawn.
We’re now no longer just accelerating growth, we’re beginning to scale.
Thirdly, growth breaks things. So as things break across all of our systems, we need to have a strong management framework in place to rapidly upgrade them so we can keep up with that scale. DAL 2.5+ needs to become DAL 3.0 needs to become DAL 3.5+. And the equivalent progression needs to happen for our Client Web and Mobile App, our internal dashboards, and all systems at the company. This is “improving the foundations.”
Fourthly, we need to “extend the moat.” This will partly come through increasing our investments in user experience, network effects, process upgrades, new process design, and premium service lines.
Every user interaction needs to be carefully mapped, reflected on, and improved. From the website through purchase to onboarding to initial delegation to multi-delegation to multi-user teams to enterprise-wide expansion, to referrals and testimonials. Everything should be magical, efficient and effective. Similarly, every process needs to be carefully thought about, reflected on, and improved.
Special attention needs to be paid to both single-player and multi-player network effects, to make sure that the more one client uses it, the better it gets for that client, and to make sure that the more clients use it on a team or at a company, the better it gets for that whole team or company, and to make sure that the more companies use it, the better it gets for all companies.
And lastly, we need to have strong leaders in place to make our Specialist Line and Strategist Line success stories. The Process Line is our core business, and will be beginning to scale. It will be difficult for these two new lines to keep up with the core business. They’ll need to compete for management attention, but with the right leaders in charge of designing them, scaling them, and upgrading them — they will produce a marvelous network effect with the core business. Firstly, we’ll have a way to retain and promote our most talented agents into more lucrative roles. Secondly, we’ll have a way to sell solutions and packages, instead of just processes. And thirdly, we’ll have a way to increase our top-line revenue with two large additional businesses. If we get this right, it will be even more difficult to compete with us.
If we’ve successfully done all of these things in Chapter 6 — beginning to scale, increasing the foundations, extending the moat — we should be able to raise up to $100M by Q4 2020.
In Q1 2020, we’ll review the foundations, once more. We’ll review the entire partnership, and make sure the right people are in the right roles, and we’re not missing key skills or experience at any level. We’ll review the management team, and make sure the team around the CEO has experience scaling their areas. We’ll review our analytics, evaluations and incentives systems, to make sure that HR and Finance have a scaleable framework to keep everyone aligned.
Then we’ll increase investment on everything on the supply side… There should be no resource limitation to driving improvement thrusts — both technology and management systems — for all platforms and systems. The Digital Assembly Line, the Client Web and Mobile App, internal dashboards, and all systems should be getting as much engineering and management time as necessary to upgrade them. Any new systems should be getting the resources they need, and then should be swiftly integrated with existing systems and swiftly upgraded.
Then we’ll add fuel. We’ll multiply spending on sales and marketing. The top-line sales and marketing goal will be to grow from between $10–30M per year to roughly 10X that. And obviously HR will need to make sure that hiring and training keep up with growth. Finance will need to continue to ensure that overhead doesn’t erode our upwards margin trajectory.
At this point, a difficult decision needs to be made about getting the management team together in a single location permanently. But at the very least, they will need to spend a significant percentage of their year in the same place.
Also at this point, the Management Team needs to become the Executive Team, as another level of hierarchy will inevitably emerge. The partnership will grow from 75 people to as many as 250. Of course, the leaner we can be, the more efficient we are. But operating in an increasingly resource unconstrained environment, we shouldn’t stint from aggressively hiring the best people for every opportunity we see that we can’t stretch our existing team to pursue.
Lastly, ironically, it isn’t until this point that I think we should actually invest in building AI. Machine learning should be driving our Process Automation by Chapter 6, already. But investing in General AI tech and other forms of deep tech to power the bot’s interactions itself begins to make sense. Up until this point, the technology investments we’ve made need to be focused on coordination, security, automation, visibility and UX. Not on magical general AI.
If we successfully complete Chapter 7, we’ll be ready to IPO. But if we’ve maintained profit discipline, we’ll be a profitable company, and won’t need to IPO. There are other ways of providing liquidity.
Investments in increasing scale, improving the foundations, and extending the moat need to continue indefinitely. But additional investments need to be made in succession planning, long-term strategy, and long-term cultural health if we want the business to survive into the next generation of leadership.
If we want to build a large, independent company that is still around in 2100, we need to think long-term about risk. To do that, we need to increase the intellectual machinery of the company. Presumably by this point the Executive Team has extensive analytics at its disposal. But Bridgewater-style investments need to be made into internal tools to surface data, identify talent, make data-driven decisions for evaluations, and to make decisions about corporate strategy.
Essentially, war-room environments need to be set up for the Executive Team and for the next layer of management beneath it. The long-term roadmap needs to be reviewed not just for technology but for every team. There needs to be deep, extensive and ongoing technology, market and user research, not just for our core business, but also for all adjacent opportunities and threats. Deep dives need to continue to occur into even the smallest details in the business, thinking about how to continue to drive deeper and deeper innovation. From these deep dives, all sorts of opportunities will emerge to extend our existing technologies and to build new ones, and we should jump at these opportunities.
Based on these war-rooms, opportunities will also emerge to launch new businesses. Because innovation outside of the core business will be distracting, I am strongly in favor of developing an aggressive spin-out program. If HR creates a strong up-or-out promotions system, then most of our best partner talent will churn. Ideally, we retain them, so we need to create an explosion of opportunities outside of the core business. One way of doing that is by aggressively launching spin-outs, where Invisible maintains control over and invests in a new startup business, and appoints one of these leaders to run it. Another way of retaining that talent is using company profits to get into the investing game, seeding alumni startups.
Amazon is the model for how to use cash in this stage of the game. Relentlessly increase efficiency, while looking for as many ways to invest in your infrastructure and invest in adjacent opportunities as possible. Steadily grow your warchest, but don’t be like Google or Apple, with no clear answer for how to use capital. Entrepreneurs should always have a good answer to how to productively use capital. So the goal is to turn the company into an innovation machine.
Let’s stop. Projecting further doesn’t make sense. As the view gets increasingly rosy, it lulls us into a false sense of security. The present risks to the business are extreme, and the business needs to become profitable as soon as possible. Until we are profitable, we are in primary existential risk. After we are profitable, existential threats will still exist, and need to be hedged, but the dominant scenario is survival.
Before concluding this exercise, it is worth re-iterating a few principles that will continue to guide the business long-term.
First, we’re committed to the principle of meritocracy. Meritocracy is painful because it creates winners and losers, which creates hard conversations, and sometimes, resentment and departures. But in the long run, it attracts and retains the best talent, and creates the most value for all stakeholders.
Second, we’re committed to the principle of transparency. We’re already making good on this by publishing as many of our internal metrics, reports and conversations as possible. There are certain conversations and certain pieces of information that should be shielded for strategic reasons, or for various privacy sensitivities. But our default will always be to share. There will be tremendous resistance to this, and this resistance will increase over time. What we must always remember is that with very few exceptions, it is dangerous to hide truth, and prudent to share it. Sharing the truth increases trust, creates alignment, increases clarity, accelerates conversations, creates shared context, cultivates public interest and invites insights and help from unforeseen sources. The most important invention of the 19th century wasn’t the steam engine or the telephone, but the modern corporation. Companies are alignment technologies, and alignment does not just emerge from incentives (see Meritocracy above), but also from beliefs, ideas, strategies, principles, values, thinking and broadly speaking, information, hence transparency. The goal is to break every record in the history of corporate transparency to become the most transparent company ever, the most public private company ever. Again, this will be hard. Partners and agents won’t like their compensation info being shared inside of the company. Investors won’t like how easily competitors will be able to research us and identify our weaknesses. They won’t like difficult decisions being shared. Some partners that are sensitive to privacy won’t like to publish. They may not be a fit here. We’re going to publish except when there’s a clear strategic or principle-driven reason to withhold.
Third, we’re committed to the principle of independence. That is why we’re obsessed with profitability. Profitability means we can’t be forced to sell. Means we aren’t at the whims of investors. Means that we can make long-term plans. But after we’re profitable, the principle will still guide us towards re-investing in the foundations, extending our moat and network effects, and generally being strategic in ways that other companies are not.
Fourth, we’re committed to the principle of expansion. Great companies are imperialistic. They dominate one market, but don’t stop at that. They expand relentlessly within that market, but also into other markets. This is why we are obsessed with designing incentives and building models that allow for continual explosion. This is why we are obsessed with finding more and more and more leverage. This is why we’re obsessed with acceleration: growth in the rate of growth, an inflection point in the inflection point. There are two types of people in the world. Clients and Future Clients.
Fifthly, we’re committed to the principle of singularity. The goal with Invisible is to build a single bot that can do everything. A single touch point for unlimited specialization. That means we need to build the ultimate Trojan Horse business. If a Client comes to us for X process, we need to learn how to do X, if we don’t already know how to do it. Once we’ve nailed X, we need to sell that client Y and Z processes, which we already know how to do well. Then we need to sell them all of our other processes, and get them to provide us with referrals and a testimonial. Then we need to get their entire team and company and network to be similarly dependent on the service. Then we need to create not only more and better processes, but more and better premium services, and continue to upsell. Eventually, we’ll build a vendor network, and there will literally be nothing you can get done through Invisible, other than the unique creative and strategic work that only you can do.
The history of business in the 20th century is different than the history of business in the 21st century: 21st century businesses need to be apocalyptic; need to think in terms of ultimates and extremes.
The reason why Apple, Google, Facebook and Amazon are all investing in AI services like Siri and Alexa is that they understand the power of being THE GATEKEEPERS of the 21st century. You cannot sell to enterprise or consumers without going through their app stores, their marketplaces, their ad networks, their platforms.
But the ultimate gatekeeper is the assistant. And the ultimate assistant has humans-in-the-loop. Is a Synthetic Intelligence, not an Artificial Intelligence. They will not invest in solving the problems that Invisible has solved and is solving until it is too late, because it is an unproven approach, because it will hurt their margins, and because they are not in the business of coordinating thousands, tens of thousands, and potentially millions of humans. Their DNA is to be high-margin, low-staff tech companies.
That’s the goal. Become the ultimate gatekeeper. Become the ultimate assistant. Give the client anything and everything they want. Do as much of it yourself as you can through your core platform. Expand into premium services to do more. But then, build an extensive vendor network to sell everything else. Mediate as much as possible, so these vendors have as little power in their relationship with the customer as possible, and so you maintain your central brokering position. With the right late-stage strategy, you can become their banker, their real estate agent, their everything. This is a position of ultimate power.
Last but not least, is the the sixth principle: human potential. This is the heart of our business. The reason our business is beautiful is not because it will become profitable and powerful. The reason our business is beautiful is that it will use technology to empower humans to live up to their full potential. It will bring technological potential and human potential into what Heidegger called “a freeing relationship”: a dialectic in which technology can discover what it is, and humans can discover what they are. This is a deflationary vision for civilization, opposed to the inflationary vision of Universal Basic Income, which is really a throwback to communism. The really futuristic thing to do is not to replace the old jobs with a stipend, but to continually replace old jobs with new jobs, to continually learn and train in more and more advanced work, and ultimately, to be free from jobs entirely: to become generals and artists, creatives and strategists. This is Nietzsche’s Superman, as I understand it. Humans are a species of becoming, we are a bridge; at our best when solving all problems, creating new ideas, imagining new possibilities, making more art, discovering more truth, building a better civilization. Imagine the most desirable possible future, and go there.
We empower — literally, give power — to partners, agents, investors and clients in different ways. There is the question of what they will do what that power. This is a question that everyone should answer for themselves, as it is the human question. But I have a strong opinion, speaking for myself as an individual, now. My hope for civilization is to sustain an exponential increase in the power of individuals and organizations, without needing to resort to using collective power. The purpose of collective power should be to do nothing other than prevent the violent use of power. All uses of power should be allowed, so long as they don’t prevent someone else from using their power freely. This is the most hopefully future that I can imagine for civilization. And it is a political position. I am not yet sure if the company should take a political position, but it will become increasingly difficult for companies with visions for ultimate technologies not to take political positions. As I am undecided on this, I will merely suggest this seventh principle — The Principle of Liberty — as a sixth guiding principle. I believe life, liberty and the pursuit of happiness will remain as compelling a vision in tomorrow’s technologically empowered future as it was over two centuries ago.