Hacker News new | past | comments | ask | show | jobs | submit login
Mattermark (YC S12) raises a $6.5M Series A to keep VCs in the know (pando.com)
46 points by dmor on Dec 16, 2014 | hide | past | web | favorite | 22 comments


First - congratulations to you and your team on the raise.

Second - a question: your last post on the topic from June 28th announcing your 2M second seed funding said that after 27 consecutive months spent pursuing some kind of funding, it was time for you to take a break from fundraising mode.

Given that was less than 6 months ago, it doesn't seem like you took much of a break after all :) Would you be willing to provide more insight and details re: what exactly changed?


Thank you.

After we announced the 2nd seed round the interest in us increased and we ended up raising more, and then when Brad Feld (at Foundry Group) offered to do the FG Angels syndicate on AngelList I figured "why not?" and I was curious to see the internal mechanics of doing one of those. So basically we raised because we've always be fans of "take money when you can".

By this point we had a ton of momentum in sales, had clarified our vision a ton in our month-long company retreat in August, and were not in a position of desperation for once. I felt like I should take heed of the old advice that "the best time to raise is when you don't need the money" so we did some analysis, had some coffees, and picked 6 firms to pitch. We committed to testing the market again.

This time, it worked.

This stuff is just messy.

she gives it away in this interview I believe: https://www.youtube.com/watch?v=UW8vbDLdlqU

in a word... "war-chest"

So is this just an aggregator of crunchbase/angellist/linkedin or some combination of scrapers and other sources?

yes basically.

i worked at a firm that used mattermark, datafox, etc. the amount of inaccurate, duplicate, and missing data was very high. all of these services are like a mashup of data sources that are not very good.

there's a reason bloomberg, thomson, etc. focus on public companies.

It is comical how often sources like crunchbase (and even the press) are wrong. Though it makes sense, private companies basically have no incentive to release any sort of data unless it plays to some narrative they are trying to create.

For example, the date the company was founded is often wrong in Crunchbase (it is reported as more recent than it actually was). Many companies raise money and don't disclose it for months (staying under the radar). Many acquisitions never publicly release the price (so the founders can claim it was a "success" regardless of whether it was).

It is just a tough space to be in to try to make something of the data that is available. Public companies are of course much easier.

Data accuracy is important but sometimes the bigger problems to solve are discovery, advanced filtering, and readability. I've played around with Mattermark a little bit and I think it does a fantastic job in solving the later problems.

From a hacker's perspective you can do curl/wget, awk, etc to aggregate and filter data from various sources yourself. But to make the same tooling available to a wider audience is what gets you to the market. Dropbox is a solid example, for instance, that made rsync super simple to use for anyone.

> Data accuracy is important but sometimes the bigger problems to solve are discovery, advanced filtering, and readability.

Discovery, advanced filtering and visualization are increasingly solved problems. Open source solutions like Elasticsearch Kibana[1] make it incredibly easy to analyze and visualize large amounts of data, and to do it better than many paid services.

For high-value commercial use cases, like those in financial services, data accuracy and completeness is all-important because your ability to identify the best opportunities and make good decisions is almost always proportional to your knowledge of the market.

Take a seemingly simple venture capital use case: I want to identify pre-Series A fintech companies in California and New York founded in the past 3 years that have raised between $100,000 and $1 million and last raised funds 4-8 months ago. If funding data and corporate information is incomplete or inaccurate, and/or funding events have not been properly categorized, the list of companies surfaced will likely exclude companies that meet the criteria, and include companies that don't meet the criteria.

For many use cases, it doesn't take many false positives or false negatives to render a data set effectively useless.

[1] http://www.elasticsearch.org/overview/kibana

agreed, we did some number crunching against crunchbase to look for the right investor profiles, making that into a usable product is much more involved

The VC industry seems like a tiny niche. But I feel like I've heard them talk about being a more general lead identification tool which has far more upside (ie, "show me all the companies in Seattle with at least 5 salespeople and at least $5m in funds raised").

A lot of the advice I have seen though is too take a niche industry and own it, then generalize. Its much easier to grab traction when you really know you exact sales targets.

The ironic thing about the niche in question is that it's arguably one that can be tackled quite happily without VC money (tech-savvy, relatively easy-to-reach B2B market, decent margin on an individual sale, small total market size but acquirer-friendly and can organically grow into other areas of finance intel driven by customer development).

Raising $11.5M in total commits them to generalising though, and there's no shortage of competent competition in the market-data aggregation space.

"The ironic thing about the niche in question is that it's arguably one that can be tackled quite happily without VC money"

At what scale and how long would that take? The VC/angel money allows you to move faster, take more chances and perhaps hire better and faster. And lose money. Important point.

Look at their pricing:


Then this:

"Today, Morrill’s curiosity and her 27-person team’s"

Imagine the payroll for those 27 people and how many subscriptions they have to sell with what is (for now) a niche product.

What are the figures on fully loaded labor costs for a 27 person team?

> What are the figures on fully loaded labor costs for a 27 person team?

Probably somewhere around $2mn - $4mn/year. That's a pretty big team for the amount of money raised to this point, they must be doing ok subscription-wise.

Somewhere (maybe the linked article) that it was 500 subscriptions for $400 a month meaning a little of 2M in annual revenue. 27 mouths to feed is a lot with that revenue.

Correct, like a better more niche discoverorg

Congratulations Danielle and team, really enjoy following along with your adventures via twitter and enjoy the daily newsletter from Mattermark. You're openness and candor is refreshing as well.

Congrats Danielle!

Congratulations! Y'all have had a wild ride so far, for being among the coolest and most practical companies in the data space

Danielle seems like a great founder.

congrats Danielle and Kevin (and team!), you guys are awesome :-)


Applications are open for YC Summer 2019

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact