I have been meaning to write about this for a minute, but the landscape report from HGL just dropped. So I am taking that as an excuse.
The Landscape Report
HGL serves to fund and incubate new political tech companies into the Democratic/progressive space. Because pretty much zero of these companies are ever going to be wildly profitable, you really need a nontraditional funder to make them possible. It’s just not a space to enter if you want to get rich. Being a tech-focused incubator gives them a good overview of the space, but also some weird incentives. One of those incentives is to not make anyone mad. This leads to their landscape report chart being *massive* and including *every single organization* that calls itself a tech company and might yell at them for being excluded. This is not a chart that will help you much if you’re trying to select a vendor.
I mean, jesus christ.
For one example that is very in my lane, here’s what they’ve highlighted for “Polling”
Again, this is weird because they’re selecting “tech companies” who do polling. What does that mean? Ehhhhhhh, vibes and online tools? Lucid is a panel vendor (or well, a panel aggregator, sort of) where you can source survey responses. They provide a marketplace for people who have survey panels and people who want survey panels. Is that a tech company? Yes, probably. Is it polling in the sense of a vendor you, a campaign, might turn to? I sure hope not, that’s a great way to get extremely wonky results. Let the experts handle that.
The other weird incentive you might worry about here is to include their own companies. I ran the chart through Claude which suggests 148 distinct companies included in this chart (also, HGL, I am BEGGING you to include a LIST. Please, for my sanity). They have a portfolio of companies here, which I also scraped copied out by hand, ughhh. My sketchy matching suggests about 30 of the included companies on their landscape report are their own. Honestly, seems low, since they have a portfolio of 60ish. This could be my lousy matching or it could be them legit not being included- but at a glance, it doesn’t seem like this list is overwhelmingly their own.
Anyway! That aside, is there anything we can pull out of this report? Specifically, from the bigass chart?
There’s So Many Companies
A hundred and forty eight companies isn’t, in the abstract, a ridiculous amount of companies. There are a lot of elections and they cost a ton of money. Where this starts to get overwhelming is that this isn’t “all the companies in Democratic politics”, this is “most of the tech-y companies in Democratic politics”. You’re missing piles of more traditional firms, consultants, various vendors, even normal things like “where you get your mail printed”. This is a relatively small slice of the industry, and it’s super crowded.
Zooming in on one category, you can start to see just how populated this is.
Content testing is acting as a kind of broad category for like, things you can give ads or text or content to and they’ll tell you what voters think, how they react, how much they click, etc. I happen to know that a couple of the big players here are missing from this chart, potentially because they operate on a more consulting model. But Blue Rose is unquestionably a hub for content testing (notably for Future Forward), as is the newer Tavern Research. This also seem to be including more digital content tracking/optimization stuff, which I would personally have split out, but I guess the chart was already gigantic.
(also sidenote, I have no idea what the company represented by the red S is. My kingdom for A LIST. Voice 123 also seems to be a voiceover marketplace? Is that content optimization or testing? Not a damn clue.)
This is a lot of options! Do we like....need all of this? Are they all serving truly different niches? Are they all *any good*? I have absolutely no idea and this report will not enlighten me, because they are not in the business of criticism. In a relatively niche industry, when trying to figure out tech with intense time constraints and little knowledge of the landscape, it’s a total nightmare to parse through all this. You need something right now immediately, and if you’re on a campaign, you’re going to be out of a job in a year and not exactly in a position to tell the vendor that they’ve lost a client if they sucked.
I know it would require being mean to people, but there’s a giant gap for recording experiences with these pile of companies and providing real recommendations to campaigns and orgs. Even the most basic of information (“Hey, do these guys actually know how to intake anything but Excel sheets?”) would be super helpful.
The Chuuuuuurn
In their 2022 report, there were only slightly fewer companies than in 2024, around 120 listed. Again, guessing at matching here, there’s 87 companies new in 2024, and 75ish who dropped off from 2022. That’s, approximately, over half of the set of companies churning in just 2 years. In a report being put out by an incubator with a strong interest in startups, I’m not shocked that there’s churn, but it highlights how tough it is to keep up on things even when you work in this industry. Every single time you go looking for, say, a content testing tool, you have to parse through “okay, oh god, what still exists, does anyone I know still work there, did they decide to pivot to something garbage or are they still good??”
Every goddamn cycle I stand atop Democratic politics asking us to stop making so many new things and then I go take a job at a new org. We are all hypocrites.
Running a tech company in politics is an incredibly tough thing. You have massive spikes in business every 4 years, a smaller spike at 2 years, and then depending on your niche it can basically evaporate between that. If you want to do a good job serving the industry, you need all sorts of knowledge and contacts not present at your bog standard startup. Being a really great political tool frequently means being a kind of lousy tool for non political use cases, or deliberately cutting yourself off from those use cases for ideological reasons. If you somehow DO get really popular with corporate users, people start getting suspicious of you not sticking to your roots and not really serving Democratic politics.
This is also, as I said above, an absolutely LOUSY place to make a pile of money. Yes, there’s a ton of money in politics, but much of it is in ads. If you’re a TV buyer, you can get a cut of that, otherwise you’re out of luck. Clients are going to bring you in late, they’re going to angst about the cost, and you can’t be 100% sure they’ll pay you before they shut down.
So, it make sense that there’s massive churn.
In my most optimistic moments, I see this as a normal filtering process that selects out the best companies. Hypothetically, if there are 5 pretty okay companies doing a thing and one really good one, 5 will die and that 1 will live, and then next election cycle there will be 1. I do realize that there’s no way that filtering is on purely merit grounds- it’s connections, initial investment, sheer dumb luck, etc. But maybe, maybe it leaves us with at-least-okay options? Hopefully?
Scattershot Diversification
My hope for this wild profusion of options for every single thing you might want a tech company to do is that it teaches the field at large to be less precious about their tools. My least favorite part of the endless VAN replacement discussions is the insistence that campaigns need something that is exactly, specifically like VAN. What they need is *a tool to track voter contacts* and *a tool to track canvassing*. This is a slight variation on a traditional CRM. If you want to sync those contacts together and run analysis on them, that’s a database and some matching software. These speciality tools are nice to have (and god knows, no one should be DIYing an app if they don’t have to), but we shouldn't feel tied to individual tools.
As large databases with pretty okay web interfaces have gotten cheaper and easier, the field has started moving in that direction. This can really free you up from the hold of companies that have previously acted as an interface layer on top of a shittier database with worse tools. Now, with good self service stuff like BigQuery, you can more easily DIY more complicated tools. This has been a huge shift for the data analytics side of the field, and imo a good one.
What I want for the rest of the tech field is to prioritize that same platform neutrality and interoperability. If folks who bounce around from orgs to campaigns are more comfortable with a variety of tools, they can be more flexible about what they use (and incidentally, have better transferable skills! It is really not a good practice to make everyone in your industry-that-fires-you-every-year learn software tricks they will use nowhere else!).
If you gave me a magic hammer, I would force every single company that wants to operate in Democratic tech to ensure that their hot new thing has built in transfer capabilities to avoid locking customers into one tech stack. Your sweet new AI email writer had better let me export all my work. Your chat tool for my volunteers ought to give me their contact info and history as a CSV. This year, the folks at MIG spent a *giant* pile of money making tools to move data in and out of VAN with intelligent retrying when it inevitably failed. That is a wild use of money. Please let us never, ever have to do that again. (And, similarly magic hammer thoughts, if you have to spend that much money propping up a tool, you should switch tools and spend that money on training.)
With my second hypothetical magic hammer, I would make everyone working in the field using this tech spend like a week look at the normal commercially available tools that do whatever their niche tool does. It’s healthy for us all to remember that the entire rest of the world also needs many of these functions, and has ways of handling them. I think you can get really locked into a way of working if you’re moving around with Dem politics, and it’s useful to briefly break out of that. You should be able to expect that your tools aren’t going to randomly go down, and that tech support will exist, and that they will talk to other tools cleanly. These are all reasonable things to want, and if you spend too long hacking solutions into place around whatever pretty-okay thing you’re stuck using, you can forget that. Please no stockholm syndrome for shitty tools.