Nine Years, No Pivot: Why the Thesis Didn't Change
"No pivot" is a claim that earns nothing unless it is made precisely, so here is the precise version. Between a September 2017 email and the 2026 network, the thesis never changed: people should hold and control their own identity and data, with cryptographic proof standing where platform trust stands today.
The thesis, in the letter's own words
The 2017 letter, documented on its own page, put it this way, translated from the Turkish:
One universal username and password instead of dozens. Granular control over which service can access which piece of personal data.
and:
In a century where data is more valuable than oil, the fact that we, the people who produce that data every moment, have no say over it, is absurd.
Strip the 2017 phrasing ("username and password" for what would become keys and credentials) and that is, recognizably, the company's present-day position: identity as something the person carries and selectively discloses, not something each platform re-collects and owns. The wedge Solidus bets on today, credential portability, is the same sentence with nine more years of vocabulary.
What was cut, and why that is not a pivot
The 2017 letter carried four threads: the identity layer, distributed compute and storage rental, rewards for user-generated content, and peer-to-peer bandwidth sharing. Three of the four are gone from scope. Solidus builds the identity thread; the validator network that exists is distributed for consensus, not general compute; the content-reward and bandwidth threads survive only as documented past consideration.
A pivot is a change of direction, the thesis replaced by a different one when the first stopped being believed. This was a narrowing: three siblings cut so the load-bearing thread could be built at all, by a team that was never large enough to build four. The distinction matters because the two failure modes it separates are different. Companies that pivot serially usually never believed anything; companies that never narrow usually ship nothing. The honest description of this record is: one belief, held; three ambitions, cut; and the cut is what made the belief buildable.
One more thing stayed constant, and it is the least flattering constant available: the problem the letter admitted it could not solve is still unsolved. A reward mechanism that would pay people for the use of their own data, the letter's "frankly, I haven't been able to settle anything on this", remains open in the 2026 stack. Nine years of thesis-holding did not crack it. It is recorded here because a page about consistency that only lists the convenient constants would be marketing, and this is a record.
Why nine years, and why only now
The thesis could not have shipped in 2017, and not for lack of trying harder: the standards that make user-held identity interoperable (Verifiable Credentials (2019), DID Core (2022), the selective-disclosure formats of the early 2020s, eIDAS 2.0 (2024)) did not exist yet. The arc page documents that timeline with public records. The consequence is the company's whole design posture: by the time building was possible, the pieces were other people's published standards, so Solidus composes them and says so, term by term.
What holding a thesis does not prove
uPort held a thesis. Civic held a thesis. Sovrin and Evernym held theirs, with more funding and more traction than Solidus has today, and they are the graveyard this category walks past every morning. Nine years of conviction is a biographical fact, not an argument, persistence proves commitment, and commitment has never once made an architecture correct or a market real.
Nine years bought receipts, not vindication. The receipts are the point.