Accelerator Partner Program Learnings
Kevin Hao, Partnerships
02/07/2019
Executive Summary
Partnerships revamped the original Accelerator Partner Program, increasing the amount of
complimentary technology offered to accelerator startups, and most importantly, removing
non-commercial use terms. Wolfram|Alpha API queries were increased from 5k queries/year to
4M queries/year, Wolfram|One now comes with 1M credits/year, and 1k Wolfram
Engines/year and 2 Wolfram EPCs were added to the offerings [Figure 1].
Figure 1 Prospective partners were shown a deck which included this page highlighting the technology being
offered.
The resulting program has grown the number of accelerator partners from 4 to 10 partners. So
far, Partnerships has seen an increase from 0 to 3 startups who have expressed interest in
utilizing the technology being offered.
It can also be seen that the accelerator space is dominated by five major players who together
make up nearly 60% of all US startup accelerator exits. The remaining US-based accelerators do
not have much of a meaningful impact. In particular, university-based accelerators are generally
poorly staffed and operated. In future implementations of the program, it is recommended that
partnerships directly reach out to startups that the dominant five accelerators invest in.
1
Strategy
The predecessor Accelerator Partner Program had two main problems: (1) accelerator startups
were not using the offered technology, and (2) Wolfram is experiencing difficulty capturing
startups that are not already aware of our technology.
There are three potential reasons accelerator startups were not utilizing the predecessor
program: (1) while in an accelerator, startups must prove product-market fit and justify
incorporating new features and technologies through customer activity. The predecessor
program forced startups to make an investment just to experiment with their business model.
(2) Alternatively, accelerator management may not be informing startups that Wolfram tech is
available to them. (3) Lastly, the current amount of technology offered is too insignificant to
perk interest from any startups.
F500 technology corporations typically offer an amount of technology in the $100k price range,
and sponsor select accelerators. This can be seen through the cohort benefits of three well-
known accelerators: Capital Factory, Y Combinator, and TechStars. Capital Factory startups
receive more than $250k in potential total hosting credits from Amazon, Google, Microsoft,
IBM, and others that last most startups 1 to 2 years. Y Combinator startups receive more than
$100k free hosting from major cloud hosting companies. Startups in the Techstars network are
granted access to over 300 perks valued at over $1M from corporate partners and sponsors,
mentor companies, Techstars companies, and others.
To address these issues, the revamped program: (1) removes red tape from the accelerator
package by granting startups commercial-use permission while part of the accelerator, (2)
significantly increases the amount of technology offered, and (3) creates PR buzz to ensure
startups are seeing Wolfram’s offerings, as well as incentivize new accelerators to partner.
From a marketing perspective, the revamped program sweetens the deal, allowing startups to
experiment with our technology and justify further investment into Wolfram technology. The
overall messaging to startups is warmer; rather than forcing startups to spend development
time and money to try Wolfram technology, Wolfram is now opening the door for startups to
do whatever they need to do to grow.
These changes hold little to no inherent risk for Wolfram. Startups in the accelerator phase
typically do not have the financial resources to afford Wolfram technology as is, and are
therefore unlikely sales targets for Wolfram. Should a startup see sudden success while at an
accelerator, the offerings still have ceilings to prevent abuse of the program. The program ends
once a startup exits the partnered accelerator. The startup should now have justified a product-
market fit with Wolfram technology as part of the solution, and become a potential customer.
2
Furthermore, with the exception of the Wolfram|Alpha API, none of the offered products
impose substantial monetary expenses on Wolfram’s part.
Implementation
Partnerships had discussions regarding the revamped program with 34 accelerators, four of
which were affiliated with the previous program. All four have adjusted to the new program,
and six new accelerators have joined the program. Partnerships is still in open discussion with
three accelerators.
When selecting accelerators for the program, partnerships preferred accelerators that were
based in the United States, had been operating for at least 2 years, had organized cohort cycles,
and focused on startups developing software.
While Partnerships was unable to execute a full press release, Wolfram released an
announcement during the Wolfram Technology Conference 2019 for the revamped accelerator
partner program. Partnerships has seen one accelerator reach out to us through that
announcement so far. Additionally, one partner, RWV Studios, selected Wolfram as one of
three inaugural technology partners in a press release on their website and social media.
Learnings
The majority of large universities have an accelerator, but these accelerators are
typically disorganized and poorly staffed.
Of the 34 accelerators that Wolfram had discussions with, 13 accelerators were affiliated with a
university. Four of these accelerators joined the program. The main roadblock was a lack of
full-time personnel running the program. Three university-affiliated accelerators were
particularly difficult to sign with, as the university system had many stakeholders throughout.
The points of contacts at these accelerators had trouble moving the conversation to a decision-
maker who could sign an MOU.
Many accelerators expect sponsorship/money from organizations they are
partnered with.
Three accelerators, Polsky Incubator (University of Chicago), Skydeck (UC Berkeley), and
Communitech were at first open to discussing usage of Wolfram technology within their
cohorts. However, upon finding that the program did not offer any sponsorship, all three
incubators declined to join the program.
3
The vast majority of accelerator-related startup successes come from a small
number of accelerators.
To measure accelerator success,
metrics such as number of exits
and number of investments can be
observed. After some data
cleaning, Crunchbase holds
information on 834 US-based
accelerators. Only 141/834
accelerators had at least one
successful startup exit. A successful
exit is defined by an IPO or
acquisition.
Considering the total number of
exits as the metric for accelerator
success, 1600 exits have been
recorded, of which 955 (59.7%)
belong to five accelerators:
Techstars (19.8%), Y Combinator Figure 2 Crunchbase holds information on 1600 accelerator
exits, of which 955 (59.7%) belong to five accelerators:
(17.1%), 500 Startups (13.3%), Techstars (19.8%), Y Combinator (17.1%), 500 Startups (13.3%),
MassChallenge (5.1%), and Plug MassChallenge (5.1%), and Plug and Play (4.4%).
and Play (4.4%) [Figure 2]. All other
accelerators each encompassed 2% or less of the total exits. These aforementioned five
accelerators stand out among the rest. The 75 th percentile number of startup exits is 6.5
exits/accelerator.
Alternatively, the total number of investments can be considered as an accelerator success
metric. The 75th percentile number of startup investments made is 27 investments/accelerator.
Considering the number of investments made, it is observed that the same five accelerators
stand out (with the addition of accelerator SOSV) [Figure 3].
Another way of measuring
accelerator success is
considering the percentage of
investments made result in a
successful exit [Figure 4].
4
Considering only accelerators that make up this 75th percentile of exits and investments, there
are 34 remaining organizations. The average success rate using this measurement is 10.5%. The
34th most successful accelerator, by this metric, is SOSV, which has an exit rate of 2.4%.
Considering the original total count of 834 US accelerators, it is obvious that effort is best
exercised on the top 34 or so accelerators.
Why is there such a disparity in the accelerator space? One potential explanation would be the
amount of funding provided to the accelerator from sponsors and corporations. With a limited
amount of funds that can be dedicated to working with startups, sponsors must be selective
with which accelerators they partner with. As a result of the disparity in available funding, the
quality and number of startups applying to each accelerator is antipodal. Lastly, the strength of
the startup application review process, accelerator personnel team, and available resources are
likely much higher for accelerators with larger available funds.
Figure 4 The percentage of exits/investments made is considered as a success metric for accelerators.
Shown are the top 34 US-based accelerators, based on this metric. The average rate of success is 10.5%,
with the 34th most successful accelerator holding an exit/investment rate of 2.4%.
5
Recommendations
Wolfram should focus on commercial accelerators over university accelerators.
Also worth noting is that few of these top 34 accelerators are university affiliated – StartX
(Stanford) and Berkeley SkyDeck (UC Berkeley) are two exceptions [Figure 4]. The majority of
the accelerators partnerships has reached out to so far are university affiliated.
While most large universities have some sort of accelerator, and it is easy to make a large list of
such accelerators and reach out to them, the pattern seen so far with university affiliated
accelerators is a lack of responsiveness and general disorganization on the part of their staff. As
a result, most such accelerators do not have a high number of exits and lack a solid vetting
system for startups interested in joining their cohorts. Sensibly, university accelerators are built
with the primary focus of fostering an interest in entrepreneurship within the student body,
rather than trying to identify key startups and bring them to success.
Students at university accelerators interested in utilizing our technology should instead be
converted to sales leads, as sales can use this to push for site-wide university licenses.
It is necessary to establish even deeper relationships with a highly selective
group of accelerators in order to increase startup technology usage.
The predecessor accelerator partner program saw zero technology usage from any startups. So
far, with the implementation of the new program, three startups have expressed interest in
utilizing Wolfram technology: Well Principled (Y Combinator), propmix.io (Innovation
Incubator), Simbiosys (Enterpriseworks at UIUC).
While this is certainly an improvement from zero startups, Wolfram should seek to deepen
relationships with a more selective group of accelerators by directly reaching out to startups to
ensure that they are aware of the advantages in using Wolfram technology.
Partnerships should browse the cohorts of major accelerators for startups that
are most likely to see improvements from using Wolfram technology.
While certain accelerators will continue to refuse to join even the revamped accelerator
partner program without sponsorship from Wolfram, if their startups are interested in using
Wolfram technology, the accelerators may agree to negotiate terms with Wolfram. After all, the
primary goal of commercial accelerators is to see as many of their startup investments to
successful exits as possible.
In regards to next steps, Partnerships should observe startups that successful accelerators have
invested in and attempt to establish direct relationships with those teams. Y Combinator,
6
Techstars, 500 Startups, Mass Challenge, and Plug and Play all publicly release the names of
startups they have invested in. Should any of those startups show interest in using Wolfram
technology, Partnerships can then leverage the relationship to sign the accelerator to the
partner program, or directly propose a deal with the startup.
Accelerator List
* = Inbound Lead
Previously Signed Accelerators
- Capital Factory
- CSI Tech Incubator (CUNY Staten Island)
- MIT Enterprise Forum Greece (MIT)
- LearnLaunchx
Newly Signed Accelerators
- RWVStudios *
- HAG Venture *
- Y Combinator
- CUNY Startups Accelerator (CUNY)
- iVenture Accelerator (UIUC) *
- Innovation Incubator
Declined to Partner
- Communitech
- Skydeck (UC Berkeley)
- Polsky Incubator (University of Chicago)
Continuing Discussions
7
- MIT Enterprise Forum Arab (MIT)
- Pioneer.app *
- Enterpriseworks (UIUC)
No Response
- All Turtles
- Afwerx
- Collaborizm
- Hatch Pitch
- ISU Accelerator (Illinois State University)
- Ignite U NY
- Precellerator
- Innovate518 (University of Albany)
- MIT Enterprise Forum NYC (MIT)
- Free Ventures (UC Berkeley)
- Entrepreneurs Roundtable
- Plug and Play
- nFactorial
- Capital Innovators
- Idea (Northeastern University)
- iStart Valley
- Chris Pfraff Media
- Bunker Labs
