Corporate Accelerator Model

This is a high-level, sanitized version of a plan I wrote for a company that wanted to start an accelerator. It’s partially from a corporate and hardware perspective. Parts have been redacted. I hope that it helps you in your own accelerator development. I’ve started and operated three startup accelerators so this is based on my perspective.

1. Proposed Plan – Starting Point

I wrote this based on my experience operating three very different startup accelerators on three continents and from assisting other programs. This is not a research paper. It’s my starting point for [company’s] situation based on what I have learned from our meeting.

Given the above and our meetings, a high-level draft plan from idea stage to operations is presented below. I have taken the approach of filling in some of the document based on my experience and current knowledge of [company’s] goals. This is the start of the discussion, research, and program building.

1.1      Establish Goals and Program Vision

1.1.1      Discussions with involved stakeholders on program options. Formulate goals for program.

1.1.2      Decide on program type relevant for [company]. While other accelerator models include the General, Hardware, Industry-specific, Toolkit-focused, Studio, Corporate, Challenge-based, University, and other types of programs, [company] is doing something different. Have clear program goals in order to construct the program.

1.1.3      Determine how to attract companies and founders in related industries: semiconductors; big data and AI; next-gen displays, AR/VR; energy; digital manufacturing; life sciences, other materials engineering.

1.1.4      Set program lifespan. Assign budget to keep program operational for a minimum number of years (accelerator strength comes partly from program continuity).

1.1.5      Set company investments. What type of financial investment will be made in portfolio companies? Again, there are many options including notes/SAFEs, direct equity investment, equity without investment, future option to invest. What is the best fit for [company]?

1.1.6      Set funding. How is the program funded? Is there a reserve for potential follow-on funding?

1.1.7      Describe and start supporting research. Gather input from partners participating in the program. Perform market research on other comparable examples. Interview customer partners and industry groups.

1.1.8      Determine KPIs. There are no industry-wide accepted KPIs for corporate or other accelerators. In each case corporate accelerators should generate their own KPIs based on what is important to them individually.

KPIs should be developed through an internal process. Relevant KPIs might include financial metrics around market size, potential to help company exit, exit multiples, and more.

1.2      Multiple Founder Models

Founders attracted to a deep tech-focused program can be different than those in a general startup accelerator. Strong candidates in [company’s] program may not want to run their own company (they may be strong researchers but weak entrepreneurs) and they may not know that their capabilities are related to the semiconductor industry. Therefore, accept multiple founder types and work with them differently.

1.2.1      Entrepreneurial founders. Treat these founders as you would other early-stage founders.

1.2.2      Research founders. Few accelerators are successful with research-centric applicants. Related to that, many universities, for example, find it difficult to commercialize technology from their funded labs because excellent researchers are not always excellent entrepreneurs. [company’s] accelerator can benefit by welcoming in research founders differently.

1.2.3      Adjacent founders. As above, pursue those not directly in industry (or who do not think of themselves in industry,

[company] can bring in management teams (internal or external) to handle the business side for research founders. A model for adding management talent is seen in studio accelerators today.

Alternatively, some founders may want to join [company] directly. This is an option to consider where the research itself is of greater value within [company] than outside.

1.3      Avoid likely failure points

Part of improving success is removing likely points of failure. I’ve noticed commonalities in what founders are bad at doing or should not be doing. These failure points are also relatively easy to avoid if identified up front. Remove these likely failure points by doing them for the founders or mandating that they have a solution.

• Founder agreements (just enforcing that one item reduced my programs’ failure rates by 20%).
• Employee agreements.
• Proper corporate entity formation.
• Grant writing support.
• Recruiting support.
• Professional website, logo, and other materials.
• Help with finance and accounting, messaging, operations.
• Professional online presence. Professional bios and photos, etc.

1.4      Staff and Support

Typical accelerator staff and support includes the below list. Managing the staff and support is often challenging. (Experienced people don’t always add value in an accelerator environment.)

• Program director (runs programs, startup advisor, holds office hours, connects teams to investors and talent, builds program awareness).
• Operations (manages mentor schedules, investor visits, workspace, housing/travel, events).
• In-house developers and designers (to aid cohort companies as needed).
• Legal services, patent services.
• Providing access to senior decision-makers / corporate connections.
• Managing training and access to equipment.
• Recruiting and management development.
• Marketing (builds program awareness, helps companies with their marketing).
• Mentor list (can number from the tens to the hundreds, across various backgrounds). Many mentors engage on a voluntary basis. Few are paid or offered equity.
• Support for visa applications, business bank account setup, incorporation, accommodation, post-program support.

1.5      Access to Equipment

Data to be collected and perspective set.

1.6      Curriculum

Data to be collected and perspective set.

1.7      Mentors / Advisors

I’ve polled many accelerators on the way they use mentors and advisors in their programs. There are two main schools of thought. Determine [company’s] style as part of discovery process.

Opt-in: All mentor meetings are optional. These accelerators want their startup founders to do what they believe is best (including on whether to meet mentors or not) rather than follow orders.

Mandatory: Mentor meetings are mandatory and frequent. In extreme cases, accelerators set up founders with 100 to 200 mandatory mentor meetings in the first month of program involvement. The purpose being to break down founder preconceptions, intentionally let them hear hundreds of opposing views, get connected to potential advisors, customers, and investors.

1.8      Investors

The investment community should see the accelerator as a source of strong deal flow. Investors should be happy to be invited to meet companies, attend demos, and more. Set expected behavior for the way invited investors treat portfolio founders. Develop relevant investor list.

1.9      Demo Day

A good demo day ends with most of the cohort on their way to receiving funding. This means that apart from packing the demo day event with the right attendees, the accelerator has introduced relevant investors to chosen startups, the accelerator sets guidelines for investors and founders, and the accelerator guides the startups through the entire process.

Not every program benefits from a demo day. Determine if [company’s] program would benefit from a demo day or another format.

1.10   Community Building

Encourage networking via periodic events for current and former portfolio companies. I’ve found that something as simple as that online community increases success and keeps founders involved long term.

1.11   Budget Allocation

Data to be collected and perspective set.

1.12   Plan Implementation

For a more complex program, as this one is, a longer multipart rollout makes sense.

1.12.1   Setup: Gain Awareness

Generate upcoming program awareness through company network, news articles, industry publications, outreach to universities, individuals via online profiles and published research, ads, application platforms.

1.12.2   Phase I: Pre-Accelerator launch

To establish the new program and provide time to build it, first host a series of related events, possibly arranged around company themes of advanced display, advanced manufacturing, advanced materials, AI and big data, energy technologies, life sciences, semiconductor technology, etc. For example:

• Conferences where targeted talent attends seminars and meets [company] representatives and others.
• Produce a video series where company researchers talk about their work. Show the facility.
• Build a contact list of relevant university departments, startups, and industry groups internationally. In the past, I kickstarted new programs via direct outreach.

1.12.3   Phase II: First accelerator cohorts

Team selection. Look for coachability, commitment, capabilities, impact on community.

1.12.4   Phase III: Long-run accelerator

Data to be collected and perspective set.

The following is a response to questions from [company] and a plan for exploring and building a semiconductor industry accelerator program. I wrote this following our February 12, 2019 CTO office meeting and my group presentation.

2. [Company’s] questions

2.1      What are the key metrics used by other corporate accelerators?

Key metrics vary wildly in accelerators of any type. When used, key metrics tend to be different based on timing:

Early metrics: awareness, applicant volume, applicant quality, acceptance rate.

Mid-stage metrics: investment size / percent of portfolio raising capital during or shortly after program, founder referrals, percent of cohort selling into the sponsoring corporation, percent of cohort acquired by the sponsoring corporation, grants issued, and above list.

Later-stage metrics: portfolio valuation, company survival rates, investor feedback, company revenue, exits, ROI, and above list.

It is important to remember that program metrics will not be purely quantitative. In the beginning (first couple years or portfolio under 25 companies) performance is often erratic.

2.2      How can we be sure to monitor the progress and performance of the accelerator?

It’s common to set goals for each company during their program involvement. “Standard” non R&D software-centric accelerators often set growth related goals, for example on weekly customer growth. This is an example of business metrics and does not fit with engineering metrics. We should develop a unique set of metrics.

2.3      What are the legal implications to [company] around liabilities and IP?

In past programs I’ve operated we have issued a program agreement to the companies we invest in, noting what investment, terms, and services are provided, as well as the accelerator’s limitation of liability, indemnification, right to terminate the relationship, etc.

Legal implications depend on the structure of the program. Other general options include:

• Joint Development Agreements.
• [Company] forms a separate entity to invest in the startups and hold equity. Startup founders continue to own their IP. [Company] does not operate the portfolio companies.
• If startups are responding to a challenge, they may be rewarded financially for their IP (challenge prize, acquisition, licensing, etc).
• Legal agreement and training requirements before founders and team employees can access equipment, software, and other resources.
• [Company] may desire for certain research to be submitted to academic journals, or to protect IP through patenting. Portfolio companies typically continue to own their own IP.

2.4      What sort of peripheral support do corporate accelerators require (legal, HR, finance, etc)?

The following roles will be helpful in any program. Also see section 1.4.

• Legal. IP review, patent protection, patent expansion.
• HR. Both for managing personnel issues, talent acquisition for incubated teams.
• Finance. This is typically a weak point for many early-stage tech teams.
• Marketing. New programs require months of lead time to generate awareness and source potential applicants. This can be achieved through targeted press, direct outreach to known startups, universities, relevant PhD programs, and via ads.
• Application review. Review applications, interview team members, form balanced cohorts.
• Business development.
• Next round funding preparation.
• Program management.

2.5      How much does it cost to run an accelerator?

To answer this question we first need to determine goals and the type of program. The following are the major costs to run an accelerator.

• Startup funding, if included. Current benchmark for software and hardware accelerators is approximately $100,000 per company. [Company] may consider a different model, for example, with higher funding over long term involvement, or where funding remains below benchmark but where much of the value-add is in equipment and expert availability.
• Equipment and production costs such as tape out, shuttle service, EDA tools.
• Staff costs. Salary and equity for some participants.
• [Company] team costs.
• Cost of outside consultants. Wide range. I’ve seen outside companies charge in the millions to run a corporate program.
• Promotional costs.
• Workspace, events, travel, and housing. Some programs claw back some of their investment in the form of workspace rent. Events include cohort and mentor dinners, demo days, and investor roadshows. Some programs pay for team travel to and from the program. Same for housing. Variation depending on whether accelerator teams relocate or work remotely.

2.6      What is the best structure to take if the goal for the accelerator is to be a strong magnet for the most innovative and promising startups?

You won’t get there in one step. New programs have similar problems:

• Awareness. New programs must deal with the lack of awareness, proving their benefit to portfolio companies.
• There are many startup accelerators today, so what distinguishes this program from the others? Why should a startup choose this program over another? Why choose this program over working independently?
• Programs tend to get better over time. Programs become known, generate referrals from portfolio founders and others, demonstrate their worth as companies raise money and are acquired
• Building a successful accelerator does not happen in one step. I recommend a multi-stage rollout in this case, see section 1.12.

2.7      How long will it take to develop a plan?

The start of the plan is above in Section 1. I estimate that a fully researched plan could be developed in a couple months. Following the plan, [company] could commence kickoff activities and the multi-part program rollout. The actual timeline can be accommodated based on [company’s] needs.

In the past for smaller software-only programs, I have gone from program approval to opening the doors of the first cohort in as little as three months. That short timeline likely does not work in [company’s] situation.

3. Why Corporate Accelerators Fail

The list of top reasons for corporate accelerator failure, in rough order of commonality:

• The program was not provided enough time to succeed because of misaligned time scales (corporate needs to show results faster than the startups).
• Misaligned risks and rewards. Are multiple failures tolerated? Do incentives for success fit corporate needs?
• Inadequate program and staff resources and capabilities.
• Located where it is difficult to attract talent.
• Lack of internal support for program and startup investments.
• Lack of internal champions to help startups integrate and sell into corporate sponsor.
• The program ends up attracting startups that could not find other options (if corporate accelerator viewed as investor of last resort).
• Using expensive name-brand operators without keeping knowledge in-house.
• Investments viewed as competitors to internal projects.
• Less connected to sources of follow-on funding options for portfolio companies.
• Lack of long-term portfolio company support (program support ends at demo day or cohort graduation).
• Did not build a community between portfolio companies.
• Operating in a winner-take-all market where top talent goes to more established programs.
• Short-term budgeting / misaligned business model (program budget runs out, unable to maintain the program long enough to generate returns).
• Skewed metrics.