There is a particular kind of bad day that begins in a conference room far from Washington. A general counsel has found a sentence in a proposed rule. A founder has discovered that the government is not merely a customer but a maze. A chief executive has been asked for documents by a congressional committee. The product still works. The customers are still buying. Yet, with one memo or one hearing, the physics of the business have changed.
Invariant exists for that day - and, more profitably, for the months before it. The Washington firm advises companies on government relations, public affairs, strategic communications, crisis preparation, reputation, coalitions, and business strategy. That list can sound like a tray of professional services until you see the common denominator: somebody outside the company can alter the company’s future.
Heather Podesta founded the business in 2007 under the refreshingly literal name Heather Podesta + Partners. She had worked for Senator Bill Bradley, studied law, advised an airline trade group, handled tax and trade policy on Capitol Hill, and become a partner at Blank Rome. She understood Washington from both sides of the meeting table. She also believed there was room for substantive policy people who could communicate like normal humans.
The first product was translation
A company speaks in products, margins, customers, and quarters. Washington speaks in jurisdictions, appropriations, rulemakings, coalitions, and election calendars. The naïve response is to shout the company’s argument louder. Invariant’s response is to translate it: Who has authority? What does that person need to believe? Which objection is substantive, which is theatrical, and which is a warning that the proposal is dead?
This is why the firm’s sharpest line is also its least grandiose: the best idea does not always win in Washington. Process wins. It is an observation about power, but also about sequencing. A persuasive message delivered to the wrong committee, after the coalition has formed, is not persuasive. It is late.
“The best idea doesn’t always win in DC. Process wins here.”Invariant
The early model had a weakness. It was associated with one founder and one side of the aisle. That can work while power is stable. American power is not stable. Public reporting says the firm began hiring Republicans in 2014. In 2017, the founder’s name disappeared from the door and Invariant appeared in its place. The sequence matters: capability first, brand second.
One decade, six times the receipts
Federal lobbying disclosuresdo not include all firm revenue
By Bloomberg Government’s comparison, that was the largest percentage gain among the bigger lobbying firms it tracked over the decade. The number is striking, but the mechanism is more useful. Invariant expanded into verticals including health care, food and agriculture, energy and environment, financial services, national security, technology, competition policy, transportation, and infrastructure. Then it braided policy work with communications and crisis advice.
The product became the braid
Consider a defense-technology startup. It may need Congress to authorize a program, appropriators to fund it, an agency to buy it, reporters to understand it, and skeptical citizens to accept it. Hiring five disconnected specialists creates a sixth job: forcing them to agree. Invariant’s pitch is that the legislative strategy, reputation work, executive preparation, coalition building, and government-sales plan should start from the same map.
That helps explain a client list broad enough to look like someone shuffled the business pages: Apple, Business Roundtable, Charles Schwab, FedEx, H&R Block, McDonald’s, Marriott, Palantir, Red Bull, Salesforce, SpaceX, Toyota, Prudential, and New York Life have all been publicly named. Invariant has said it serves more than 225 clients, including nonprofits, associations, and startups that have barely made it past Series A.
They are not buying the same policy position. They are buying a similar reduction in uncertainty. The firm is paid through professional-service engagements rather than a scalable software subscription. That makes the model human, relationship-heavy, and difficult to clone quickly. It also means the work earns its keep when the value of avoiding one regulatory mistake dwarfs the advisory bill.
A culture designed for handoffs
Integrated firms often look better in a proposal than in a meeting. Specialists hoard clients. Senior people protect credit. Departments produce five strategies with matching covers and conflicting verbs. Invariant’s careers material attacks that failure directly: no silos, no egos, no credit-grabbing. The language is unusually cheeky for K Street. Its recruiting headline promises applicants the chance to “move mountains, change lives, eat shrimp.”
The joke is doing serious work. A multidisciplinary offer depends on people passing information and status across internal boundaries. If the communications adviser cannot challenge the lobbyist, or the policy expert cannot admit that the public argument is losing, the braid comes apart. Plain speech is not merely tone. It is coordination technology.
The approach has limits. It is a poor fit for a small company with no meaningful policy exposure, a one-off local matter, or an executive shopping for a famous contact instead of sustained preparation. It also cannot manufacture a viable public argument from a business objective that offers decision-makers nothing. Access can open a conversation. It cannot make the other side forget its incentives.
The playbook hiding in plain sight
In 2026, Invariant hired Jaime Cheshire, a former CIA and House Armed Services Committee official, to lead national-security government relations. The appointment fits the larger pattern. When a market becomes more entangled with public authority - artificial intelligence, defense technology, energy, trade - the firm adds people who understand both the formal process and the unwritten tempo.
The company has also begun publishing more of its method. A 2026 case study of the TAKE IT DOWN Act described six ingredients behind a rare standalone federal technology law: strong champions, timing, grassroots pressure, inclusive drafting, bipartisan and bicameral support, and a catalytic public moment. The lesson is almost comically transferable. Narrow the ask. Make it legible. Recruit people who can move it. Start before everyone else realizes the moment has arrived.
What another operator can copy
- Map the decision process before polishing the message.
- Translate your goal into the incentives of the people with authority.
- Build relationships across factions before control changes hands.
- Treat policy, reputation, and sales as connected when they share a gatekeeper.
- Prepare scenarios while the crisis is still hypothetical and cheap.
Podesta once said her first-year ambition was $1 million in revenue; the firm made $2.2 million. The point is not that the original forecast was timid. It is that she had found demand for translation in a city built on specialized dialects. Invariant’s later growth came from recognizing that the translation problem did not stop at Congress. It ran through the newsroom, the regulator, the boardroom, the customer, and back again.
Washington did, indeed, get more complicated. Invariant’s wager was to become complicated enough inside that its advice could feel simple outside. On a bad day in a distant conference room, simplicity is what the client is really buying.
Continue the conversation