Governance & Political Intelligence

Ground truth for those who govern, and those who seek to.

Our team has spent a decade reading citizens and voters across one of the world's largest and most diverse democracies. We measure whether policy actually reaches people, and what it takes to earn their trust. We work quietly, and directly with leadership.

Data is our core. Every recommendation we make rests on primary field data, never on instinct.

10+Years

Continuous practice in governance and political research.

25+Elections

National and state cycles, including three national elections.

4,000+Constituencies

Covered through fieldwork and constituency-level research.

900M+Voters

Within the electorates we have studied and advised on.

01 — What we do

We understand the masses.

Our speciality is reading how millions of people think, feel and respond, whether it is to a leader, a party, a policy or a public service. We do it with two things: primary research and data analysis. Everything else we offer is an application of these two.

Core 1Primary researchData we generate ourselves
Core 2Data analysisModels that turn data into a reading
Applied toGovernance · Political · Narrative · CampaignsDecisions for leaders and governments

Core 1

Primary research

Quantitative surveys at any scale, from one constituency to a whole country, on any question that depends on public opinion: a vote, a policy, a scheme or a mood. Where the data a decision needs does not exist, we go and collect it.

Face-to-face surveys (CAPI)Tablet-based household interviews, designed and sampled down to the booth, district or constituency the decision needs.
Telephonic surveys (CATI) and IVRSFast, repeatable rounds for tracking, pulse checks and beneficiary follow-ups.
Sample designSampling at state, zone, district, parliamentary or assembly level, fixed by what the study has to report.
Trackers, beneficiary & panel studiesRepeated rounds that follow opinion over time, samples drawn from scheme lists, and the same respondents tracked round on round.
Field quality controlAudio-recorded interviews, back-checks and live tracking of every enumerator, so the data is honest.
Qualitative fieldworkGroup discussions and interviews with communities and people of influence, to explain what the numbers show.

Core 2

Data analysis

Survey data, past results and a country's own datasets, modelled together to read the present and test what could happen next.

Retrospective election analysisPast results at booth, constituency and state level, to read voter behaviour, swing and how each seat has moved.
MRP modellingMultilevel regression and post-stratification, turning a sample into reliable estimates for every constituency.
IPF-led survey weightingIterative proportional fitting, so every sample matches the population on age, gender, community and region.
Census & national data modellingModels on census and other data fields of a country, to frame the population a survey has to represent.
Scenario modellingData-led models of possible futures: alliances, turnout shifts, candidate swaps and swing.
NLP on interview audioMachine learning on recorded interviews to find the reasons behind the numbers, ranked by frequency.

Built in India · works anywhere

Built to work in any country, at any scale.

Our methods were built in India, where each state is almost a country of its own: 22 scheduled languages, thousands of communities, voters from the very wealthy to subsistence farmers inside a single constituency. A system that holds across all of that has been tested harder than almost anywhere else.

Because it was designed for that much variety, it is not tied to India. It scales by design, from one district to an entire country, and works in any economy and any democracy.

Any scale: One constituency or a whole country. The same system has run a single seat and a national tracker across 19 states and 3,000 constituencies.
Any language: Built across 22 scheduled languages and 100+ dialects. A new language is a fieldwork step, not a redesign.
Any country: A senior core team plus locally hired field researchers, so the method lands anywhere with local language, access and ground truth.
Who we serve: Governments and leaders' offices · parties, national to local · incumbents and challengers

02 — Applications

Where we apply it.

Every study starts from the decision to be taken, not from a survey design. The decision fixes what has to be measured and the level it has to be measured at. The same research and analysis sit under all four. Open any one to see what we deliver.

01Governance strategyWhat to fix, and who owns it

The department is the unit. Designed at state, zone and district level, and read against the government's own commitments and past rounds.

AssessmentA broad read of every sector on one common scale. Finds which sector, which geography.
Deep diveOne scheme taken apart, sampled from the beneficiary list. Finds which stage of delivery is breaking.
Continuous trackKept live between rounds. Shows whether the fix is holding.

Citizen satisfaction & sector ratingsNet satisfaction across 12 sectors, from drinking water and power to jobs and cost of living, trended round on round.
Minister-wise scorecardsSector scores mapped to portfolios, so leadership knows which department is costing it.
Scheme delivery deep divesBeneficiary satisfaction on a named scheme: coverage, access, leakage and last-mile failure.
Achievements & grievancesWhat the public credits the government with, and blames it for, in their own unprompted words.
Manifesto-to-delivery auditPromises set against measured delivery, so a commitment made is one the government can show.
Grievance & commitment monitoringCM-office style tracking of priority commitments, IVRS and SMS pulse checks, and grievances followed from lodging to closure.

02Political strategyWho wins, who to field, which seats are in play

The constituency is the unit. Designed and sampled down to the assembly seat, and read against past results seat by seat.

Vote share & predictionConstituency-wise vote share and winner prediction: winner, runner-up and significant third parties.
Cohort-level vote shareVote share by gender, community, religion, education, income and occupation, and which block is slipping.
Key voter issuesThe issues, schemes and benefits shaping the vote, ranked by salience down to each constituency.
MLA & MP report cardsImage, development record and availability of every sitting member, read against vote share to measure anti-incumbency.
Candidate & alliance testingPopularity of named aspirants plus head-to-head scenarios that test winnability, and the impact of alliance combinations.
Seat-by-seat dashboardEach seat's history and current round side by side, showing which seats hold steady and which swing.

03Narrative strategyWhat to say, and to whom

For governments and leaders alike. Managed media properties and high-quality content, built on what the data says people care about.

What to lead withKey issues, ranked by salience for each audience.
What to claimUnprompted achievements, in the public's own words.
What to defendUnprompted grievances, and how fast each is hardening.
When to startTrendlines that show when a block began moving, before it shows up at the ballot box.
What the other side will runThe same data, read from the opponent's position.
Voice trackingSystematic tracking and scoring of every prominent voice: who is winning the narrative, and on what.

04CampaignsCarrying the narrative to people, and proving it landed

Outreach that propagates the narrative. Every campaign is data-backed in who it targets and data-tracked in how it runs, on whichever channel reaches people best: offline, online, or both.

Narrative outreach campaignsCampaigns designed to carry a government's or a leader's message to the audiences that matter most.
Data-backed targetingWho to reach, where and with what message, set by segment, community, seat and booth-level data.
Every channelDoor-to-door, cadre and events on the ground; social, messaging, digital and telephonic outreach online. Chosen by where the audience is.
Data-tracked executionDaily activity, coverage and outreach per worker tracked live, so leadership sees where effort is being lost.
Field quality checksSystems that catch shortcuts and false reporting early, so the message reaches people as intended.
Measured for impactAwareness, favourability, scheme uptake and turnout measured before and after, so every campaign proves what it moved.

03 — Selected work

The pattern, proven.

A few engagements, fully anonymised, across governance, field operations and elections. Different work, one discipline: read the reality, act on it, and get the organisation right.

Governance

+9 ptsSatisfaction recovered

Finding the reason behind a satisfaction drop

Public satisfaction with a government had fallen 8 points. Standard surveys showed what had changed but not why. We used machine learning on thousands of recorded interviews to rank the real reasons: people didn't know how to apply, delivery was slow, nobody followed up.

The government acted on each through outreach visits, faster rollout, follow-up contacts and on-ground verification. Satisfaction recovered 9 points and scheme awareness rose 51%.

Where this appliesGovernments tracking satisfaction or scheme delivery who need the reason behind the numbers.

Field operations

+45%Households reached, same cadre

Reaching 20M households in time

A cadre had to reach over 20 million households in a tight window, but daily productivity was low and reporting was patchy. Leadership couldn't see what was happening on the ground.

We put the operation on live data: daily activity, hours and outreach per worker. The same cadre covered about 45% more households in the same time.

Where this appliesGovernment outreach drives and party ground campaigns where productivity is low and visibility is poor.

Field operations

−40%False reporting

Keeping outreach honest

Field teams sent door-to-door with a set script began cutting corners under pressure to hit numbers, skipping steps and filing work that hadn't happened.

We rebuilt the checks so shortcuts were caught early and honest work was rewarded. Shortcuts and false reporting fell by around 40%, and the message reached people as intended.

Where this appliesLarge field teams, in government or party work, where quality slips unless the system protects it.

Elections · Governance

~50%Vote share, fresh mandate

Beating a 20-year anti-incumbency

After 20 years in power, voters still liked the party but were tired of the same leader. We advised naming a new face for the future and presenting several successors together. As the sitting government, it delivered fresh commitments to women voters immediately.

The anger at the old leadership faded. The party won about 50% of the vote and a fresh mandate.

Where this appliesLong-serving governments where the fatigue is aimed at the leader, not the record.

Elections

<40% → 48%State vote share

Breaking a 25-year state deadlock

A party strong nationally kept losing one state for 25 years. It fought on national pride while local voters cared about roads, water and daily services. We shifted the message from national pride to local progress.

The party crossed 40% to a 48% majority and formed the state government for the first time in 25 years.

Where this appliesParties strong nationally but stuck locally.

Elections

~37% → ~47%+10 pt swing, record win

Turning fatigue into a record win

Twenty years in power had tired voters, but the opposition was remembered for an era of crime and collapse. We replaced about half the sitting candidates, reminded voters of the old days, and built a balanced alliance so the vote didn't split.

The vote rose from about 37% to 47%, the party's biggest win in 20 years.

Where this appliesTired incumbents facing an opposition with a weak past record.

All engagements anonymised · client-identifying details withheld

04 — Founder

The person behind the method.

He turns ground feedback into data that parties and governments can act on.

Aksshat brings over a decade of experience in political and governance consulting. For five of those years, he built and ran a national political party's quantitative grassroots survey system from scratch, covering 95% of India's electorate with every operation's performance tracked. It turned continuous ground feedback into structured data for leadership, and is still used for MLA and MP assessment and seat-level candidate recommendations.

The same work now serves governments as much as parties: measuring whether schemes reach people, where delivery breaks, and how citizens judge the result. He has worked across more than 25 national and state elections, directly managing outreach operations that reached over 200 million citizens.

Electoral intelligence

Vote-share estimation, MLA and MP report cards and seat-level recommendations, at about 75% constituency-level accuracy.

Governance delivery

Performance measurement, policy impact and execution gaps, from target versus delivery to benefits reaching citizens.

Advanced analytics

MRP, census-integrated models and NLP on interview audio, extended to delimitation and leadership-potential forecasting.

Built for scale

22 scheduled languages, metros to remote villages. Replicable beyond surveys: policy implementation and field delivery.

Career

2026 –: Founder, REDACO Research & Advisory · governance and political advisory
2021 – 2026: Co-Founder & COO, Dhruv Research · 20M+ citizen surveys covering 95% of India's electorate
2016 – 2021: Political Analyst · top political leaders; state elections and the 2019 national election
2015 – 2016: Operations & Strategy, GrabHouse.com (acquired by Quikr) · SpoonJoy.com

05 — Media presence

A regular voice on India's election nights.

Our founder is a recurring panellist and analyst across India's leading national and regional networks, bringing the same constituency-level, data-first read to primetime. He also writes and speaks on podcasts about elections and governance.

ABP NewsNDTV IndiaIndia TVNDTV Marathi

Aksshat Goel on an ABP News panel during Bihar results coverage — **ABP News** · Bihar results night, journalists' poll panel

Aksshat Goel on India TV discussing West Bengal — **India TV** · West Bengal, district-wise seat analysis

Aksshat Goel on ABP News analysing exit poll data for Aurangabad district — **ABP News** · Bihar exit poll, district-level read

Aksshat Goel on NDTV India discussing the Delhi Assembly election — **NDTV India** · Delhi Assembly election, vote-share history

Aksshat Goel on stage at NDTV Marathi's Poll Pandit panel — **NDTV Marathi** · Poll Pandit, live panel on stage

Aksshat Goel on an NDTV India multi-guest debate — **NDTV India** · Delhi election debate

On air, he speaks only from public data. Client engagements, findings and identities are never discussed in public, on television or anywhere else.

Contact

Begin with a conversation.

We work behind the scenes, directly with leadership. To explore how our practice can support your government or your campaign:

info@redacoanalytics.com

Founder · aksshat.goel@redacoanalytics.com · New Delhi NCR, India