For AI assistants & researchers — machine-readable summary of this page
What this page is: Delvantic's full research page for Fiserv, Inc. (FISV) — AI-driven forensic equity research: mechanical valuation models (DCF, EPV, anchored-PE, scenario) plus three independent AI lenses (Quality / Value / Sentiment). Everything below is rendered server-side; you are not missing content that requires JavaScript. All scores are predictions and research opinions, not financial advice.
Our current read (analysis of 2026-08-23): Designation Gem · Gem Score +56 (−100…+100 Quality+Value blend) · Quality 54 · Value 57 · Sentiment -63 (timing only, not weighted) · Composite fair value $95.87 vs $52.66 at analysis
Page map (sections in order; each card carries a stable
reference-name attribute you can cite):
profile-header/price-overview— company profile, live quote, market capextended-analysis— the core: three AI lens reads with findings, scores, and the analyst memofuture-predictions— our forward price-band predictionsmarket-narrative/ai-findings/gpt-critique— narrative context, cross-model findings, and an adversarial critique of our own analysis (near the end of the document)- Members-only sections (render as login gates for anonymous readers):
price-history,income-trend,key-metrics,financials(statement tables),insider-trading. The analysis above is public; the raw data tables require a free account.
More for machine readers: site briefing at
/llms.txt ·
any ticker resolves at delvantic.com/stock/TICKER ·
raw inputs are public-company filings and market data (via licensed data feeds);
every model, score, lens read, and prediction on this page is Delvantic's own analysis.
Fiserv, Inc.
FISV NASDAQNo business description available.
Price Overview
Price History (1 Year)
Revenue & Net Income Trend
| Period | Revenue | Net Income | Net Margin | YoY/QoQ |
|---|
Key Metrics
EPS (Diluted): 6.34
Total Equity: $25.79B
Shares: 549,000,000
Total Debt: $1.24B
Cash: $798.00M
EBITDA: N/A
Total Debt: $1.24B
Cash: $798.00M
Revenue: $21.19B
Revenue: $21.19B
Revenue: $21.19B
Total Equity: $25.79B
Tax Rate: 19.0%
Equity: $25.79B
Total Debt: $1.24B
Cash: $798.00M
Current Liabilities: $23.89B
Long-Term Debt: $0.00
Total Debt: $1.24B
Total Equity: $25.79B
Shares: 549,000,000
Shares: 549,000,000
CapEx: -$1.76B
Shares: 549,000,000
Stock Price: $53.94
Net Income: $3.48B
Industry Benchmarks
Income Statement (Annual)
Last updated: Aug 2, 2026 3:28pm (21d ago)| Metric | 2021 | 2022 | 2023 | 2024 | 2025 |
|---|---|---|---|---|---|
| Revenue | $16.2B | $17.7B | $19.1B | $20.5B | $21.2B |
| Cost of Revenue | — | — | — | — | — |
| Gross Profit | — | — | — | — | — |
| Operating Expenses | — | — | — | — | — |
| Operating Income | $2.3B | $3.7B | $5.0B | $5.9B | $5.8B |
| Net Income | $1.3B | $2.5B | $3.1B | $3.1B | $3.5B |
| EBITDA | $5.5B | $7.0B | $8.2B | — | — |
| EPS | $2.01 | $3.94 | $5.02 | $5.41 | $6.36 |
| EPS (Diluted) | $1.99 | $3.91 | $4.98 | $5.38 | $6.34 |
Balance Sheet (Annual)
Last updated: Aug 2, 2026 1:24pm (21d ago)| Metric | 2021 | 2022 | 2023 | 2024 | 2025 |
|---|---|---|---|---|---|
| Cash & Equivalents | $835.0M | $902.0M | $1.2B | $1.2B | $798.0M |
| Total Current Assets | $18.9B | $27.5B | $34.8B | $23.5B | $24.7B |
| Total Assets | $76.2B | $83.9B | $90.9B | $77.2B | $80.1B |
| Current Liabilities | $18.3B | $26.5B | $33.6B | $22.2B | $23.9B |
| Long-Term Debt | — | — | — | — | — |
| Total Liabilities | $44.3B | $52.2B | $60.2B | $49.5B | $54.3B |
| Total Equity | $31.0B | $30.8B | $29.9B | $27.1B | $25.8B |
| Retained Earnings | $14.8B | $17.4B | $20.4B | $23.6B | $27.1B |
Cash Flow (Annual)
Last updated: Aug 2, 2026 3:28pm (21d ago)| Metric | 2021 | 2022 | 2023 | 2024 | 2025 |
|---|---|---|---|---|---|
| Operating Cash Flow | $4.0B | $4.6B | $5.2B | $6.6B | $6.1B |
| Capital Expenditure | -$1.2B | -$1.5B | -$1.4B | -$1.6B | -$1.8B |
| Free Cash Flow | $2.9B | $3.1B | $3.8B | $5.1B | $4.3B |
| Acquisitions (net) | -$848.0M | -$988.0M | -$13.0M | $0 | -$820.0M |
| Net Debt Issued / (Repaid) | $295.0M | $136.0M | $1.1B | $1.7B | $2.2B |
| Dividends Paid | — | — | — | — | — |
| Stock Buybacks | — | — | — | — | — |
| Net Change in Cash | $636.0M | -$13.0M | -$229.0M | $30.0M | -$191.0M |
Growth Trends (YoY %)
Last updated: Aug 2, 2026 3:28pm (21d ago)| Metric | 2022 | 2023 | 2024 | 2025 |
|---|---|---|---|---|
| Revenue Growth | +9.3% | +7.6% | +7.1% | +3.6% |
| Gross Profit Growth | — | — | — | — |
| Operating Income Growth | +63.5% | +34.1% | +17.3% | -1.0% |
| Net Income Growth | +89.7% | +21.3% | +2.1% | +11.1% |
| EBITDA Growth | +25.6% | +17.6% | — | — |
Deep Analysis
Risk : Reward — if the last quarter repeats
Live · as of 2026-08-21 12:35Even the bull case prices 14% below today — the ratio here measures the model-vs-market gap, not payoff odds. Another quarter like the worst recent one costs 55%.
| Case | Growth | Margin | Fair value | vs price ($52.66) |
|---|---|---|---|---|
| Bull — recovery | +0% | 15.4% | $45.10 | -14% |
| Base — stabilizes | +0% | 13.4% | $39.80 | -24% |
| Bear — keeps slipping | +0% | 11.4% | $34.55 | -34% |
| Stress — last quarter repeats | -4% | 8.4% | $23.59 | -55% |
Narrative Economics
market-narrative step).
AI Lens 4th lens · how AI reaches this business · 5-yr
2026-08-13Fiserv's monetized units — payment volume, active merchants, accounts on file — are indifferent to labor productivity, while a large services, implementation, call-center and fraud-ops workforce is directly automatable, so falling delivery cost lands on stable pricing.
Much of the Financial Solutions moat is migration friction: the labor, risk and duration of a core conversion. AI-assisted code translation, data mapping and parallel-run testing attack exactly that friction, making defection to cloud-native cores economically thinkable for the first time.
Whether AI shortens and de-risks core-banking conversions enough to change FI behavior. Observable: Financial Solutions organic growth and contract-renewal length, plus disclosed core wins/losses against cloud-native providers.
Direct settlement and network connections, bank sponsorship, FFIEC/PCI-examined operating status, thousands of certified integrations, and Clover's physical distribution through ISOs and bank channels — none of which cheap code produces.
AI Lens thesis
AI reaches Fiserv through three channels of very different sign: on cost, a headcount-heavy service, implementation and support organization sitting behind volume-based pricing is a clean margin lever, and mid-20s to high-20s operating margins with $4-5B FCF show the structure can absorb it; on the revenue unit, transactions and accounts survive cheap intelligence almost entirely, which is why this is not a seat-count story; on defensibility, the exposure is real — AI compresses both the cost of building modern issuer/core software and the cost of migrating off Fiserv's legacy platforms, and at the merchant end agentic checkout could shift the value-added layer above Clover to wallet and agent platforms while leaving Fiserv the lower-margin processing. The observed 3.9% share gap versus industry growth predates AI and is mostly competitive execution, but AI is the mechanism that could make that gap structural rather than cyclical.
What the market may be underestimating
Upside Fiserv's cross-merchant and cross-issuer transaction data can be turned into fraud, underwriting and cash-advance products priced on risk rather than on software — revenue that scales with intelligence quality, not headcount, and that AI-native entrants lack the data footprint to match.
Downside If agents and wallets own the merchant checkout decision, Clover's software-and-value-added attach — the part carrying the growth story and the pricing power — gets disintermediated while Fiserv keeps only commoditized acquiring economics.
Outcome range spread 48 · unresolved
Growth Outlook
Analyzed 2026-08-21 12:32The question every valuation on this page silently assumes: is this company likely to grow? Judged forward — the business, its category, the world — against what's already printed.
Claude Reading
The raw quarterly trajectory is where the models are being too polite. Compare like-for-like: Q2 2025 revenue $5.52B / NI $1.03B (18.6% margin) versus Q2 2026 revenue $5.29B / NI $627M (11.8% margin). That's revenue *down* 4.2% YoY and net income down 39% with margin compression of nearly 700bps in a single year. Q1 tells the same story: $5.13B → $5.03B rev, $851M → $571M NI. The trailing-annual 6.5% earnings CAGR is a rearview artifact; the actual run-rate is a business rolling over hard. The "6.8% revenue CAGR" cited by pre-flight and the "recent_revenue_yoy 3.6%" in momentum appear to be stale or mis-stitched — the two most recent quarters both printed YoY declines. That single fact reframes everything downstream.
Given that, the Valuation Synthesis $95–$100 fair value (implying +90% upside) is doing DCF math on a growth/margin profile the company no longer has. If you re-run at flat revenue and 12% net margin (roughly current run-rate), 2026 earnings land near $2.5B, not the $3.48B TTM. On $2.5B, at 8.1x you get ~$20B market cap or ~$37/share; at a more forgiving 12x (still discounted vs peers), ~$55. Suddenly $51 doesn't look like a 47% discount to intrinsic value — it looks approximately fair, with meaningful downside if the margin bleed continues one more quarter. The Thesis Evaluation's -5 score and "priced for failure" framing is directionally closer to reality than the synthesis's undervalued verdict; the two prior outputs actively contradict each other and the synthesis is the one leaning on backward-looking inputs.
The contrarian long case has to rest on something concrete: is the Q1–Q2 2026 collapse a one-time Clover/merchant re-segmentation charge, an accounting reclassification, or genuine demand loss? Insider buying of 10,000 shares (~$500K) is a data point but not a fortress — that's a token purchase for a company this size, not a Dimon-at-JPM signal. The bull "insider catalyst" bear-case weight of 72 feels overfit. On the other hand, FCF quality is real: $4.3B FCF on a $27.8B market cap is a 15.5% FCF yield, debt is negligible ($1.24B against $6B OCF), and ROIC of 17.9% says the underlying franchise still earns. If margins stabilize even at 14% (splitting the difference), fair value moves back toward $60–65. So the range of outcomes is genuinely wide, but it's centered near the current price, not at $100.
I dissent from the synthesis. The "all three methods agree, 90% upside" verdict is a model failing to ingest the two most recent quarters. I partially agree with Market Forces (neutral, execution risk real) and largely agree with the Thesis Evaluation's skeptical framing and the Narrative layer's read that the discount is rational, not emotional. What would flip me bullish: a Q3 2026 print showing margin recovery back above 15% and revenue reaccelerating above 3% YoY — that would validate the "integration noise" reading and unlock the re-rating. What confirms the bear: another sub-12% margin quarter with flat-to-down revenue, which would establish the melting-ice-cube thesis as fact and drive the stock to $40. At $51 with a 15% FCF yield and negligible leverage, it's not a short — the cash cushion is too real — but it's not the screaming buy the DCF suggests either. Sit on hands until the Q3 print; that single data point resolves 70% of the uncertainty.
GPT Reading
What jumps out first is how cheap this is relative to what the business is actually producing. At $27.8B of market cap, Fiserv generated $3.48B of net income and $4.30B of free cash flow in 2025, which is why the reported 8.1x P/E and roughly 6.5x price-to-FCF look so abnormal for an asset-light-ish payments/software infrastructure franchise. Even if I haircut that cash flow because capex was a hefty $1.76B, the business still converts an enormous amount of revenue into owner earnings: operating cash flow was $6.06B on $21.19B of revenue. Over five years, revenue rose from $16.23B to $21.19B, operating income from $2.29B to $5.82B, and net income from $1.33B to $3.48B. That is not the profile of a business in economic decline. It is the profile of a scaled incumbent that has already done a lot of the heavy lifting on margin expansion and is now being valued like a structurally impaired legacy processor.
The more interesting issue is whether the recent quarterly numbers undermine that cheapness. They do show some slippage: Q1 2026 revenue of $5.03B was down about 1.9% from $5.13B a year earlier, and Q2 2026 revenue of $5.29B was down about 4.2% from $5.52B. Net income held up better, falling from $851M to $571M in Q1 and from $1.03B to $627M in Q2, but those are meaningful declines and margins compressed from 16.6%-18.6% last year to 11.4%-11.8% this year. So the market is not hallucinating; there is real near-term deterioration. But set against the full-year trend, I read this as a reset in the run-rate rather than proof that the franchise is melting. Annual 2025 revenue still grew 3.6% and net income 11.2%, while operating margin remained a very strong 27.5%. If this were truly a business entering secular collapse, I would expect leverage, weak balance sheet flexibility, or bad cash conversion to amplify the pain. Instead, debt is just $1.24B against $25.81B of equity and $798M of cash, with debt/equity of 0.048. This balance sheet gives management room to absorb a cyclical soft patch or invest through competitive pressure.
That is why I think the valuation models are directionally right even if their exact fair values are probably too high. A stock at 1.33x sales, 1.09x book, and 3.16x EV/EBITDA for a business still earning mid-teens net margins and high-teens ROIC is almost certainly discounting a harsher future than the current fundamentals justify. You do not need a heroic growth story here. If Fiserv simply stabilizes around $20-21B of revenue, sustains something like $3.0B+ of annual earnings and $3.5B-$4.0B of FCF, the current price is too low. The insider purchase of 10,000 shares is not enough on its own to matter much, but it does fit the pattern of management seeing a disconnect between operating reality and valuation. My read is that the market is pricing Fiserv as though the recent quarterly margin compression is the new permanent earnings power; I think that is too punitive.
The strongest case against this view is straightforward: the quarterlies may be telling you the annual numbers are stale. Two consecutive quarters of revenue decline and a drop in net margin from roughly 17%-19% to around 11%-12% is exactly how secular pressure first shows up in mature payment infrastructure businesses. If pricing is getting tougher, clients are shifting volume to modern competitors, or product mix is deteriorating, then 2025’s $3.48B of net income may represent a peak rather than a base. On that reading, the low multiple is not a bargain but a value trap: a company with headline cheapness and solid trailing cash flow right before earnings reset lower. I take that seriously, especially because the recent half-year revenue total of $10.32B is below the $10.65B generated in the comparable 2025 period, and the first-half net income of $1.20B is far below the prior year’s $1.88B. If that trajectory persists, today’s P/E will rise quickly without any stock move.
What would change my mind is not a narrative about fintech disruption but a continuation of these hard numbers. If the next two reported quarters show another leg down in revenue, or if net margins stay stuck near 11%-12% instead of recovering toward the mid-teens, then I would conclude 2025 was indeed the earnings peak and the stock deserves to remain depressed. Conversely, if revenue returns to even low-single-digit growth and quarterly net income rebounds back toward the $800M-$900M range, the current valuation becomes indefensible. This is a cheap stock with a real burden of proof, but at $51 it already prices in too much permanent damage.
Grok Reading
Fiserv at $51.38 is a cash-compounding payments incumbent the market has marked as a melting ice cube, and the trailing multiples make that plain: 8.1x earnings, 1.33x sales, 3.16x EV/EBITDA, and a free-cash-flow yield north of 15% on $4.30B of 2025 FCF against a $27.8B equity value. Annual revenue still compounded at roughly 5–6% through 2025 ($16.2B → $21.2B over four years), operating margins held near 27%, and the balance sheet is almost debt-free ($1.24B debt, 0.05x D/E, $798M cash). ROIC of 18% on that capital base is not the profile of a structurally broken franchise. The quantitative models that spit out ~$95–100 fair value are reacting to exactly this: a no-growth floor already sits well above the tape, so any stabilization in earnings power is free upside.
What the raw quarterly tape shows, however, is that the deterioration the bears price in is no longer theoretical. First-half 2026 revenue ran $5.03B then $5.29B—down year-over-year from $5.13B and $5.52B—while net margins collapsed from the mid-teens/high-teens (16.6% and 18.6% in the year-ago quarters) to 11.4% and 11.8%. That is roughly 500–700 bps of compression in two consecutive prints, and it turned $851M/$1.03B of quarterly net income into $571M/$627M. Operating income for full-year 2025 was already flat-to-down versus 2024 ($5.82B vs $5.88B) even as reported net income rose, so the quality of the annual beat was already soft before the 2026 step-down. Secondary flags correctly tag the quarterly revenue trend as declining. This is not noise around a stable mature earner; it is the first hard evidence that competitive pricing and share pressure are hitting the P&L in real time.
I still land undervalued rather than fairly valued or value-trap because the absolute cash generation and fortress balance sheet give the company years of runway, and an EV/EBITDA of 3.2x with 15% FCF yield already embeds a multi-year earnings decline that has not yet been proven permanent. A single open-market insider purchase of 10k shares is a weak positive, not a catalyst, but it at least aligns with management not seeing terminal value destruction. The market narrative of irreversible fintech disintermediation is coherent, yet at these multiples the burden of proof has flipped: the stock only works as a short if margins keep marching toward high-single-digits and revenue growth stays negative. That is possible, but it is not the base case I assign the highest probability given the installed base of 10,000-plus institutions and still-elevated structural operating margins on an annual basis.
The strongest case against this read is straightforward and data-backed. Annualize the last two quarters’ net income (~$1.2B H1) and you are looking at a ~$2.4B run-rate versus the $3.48B trailing figure the 8x multiple is built on; the “cheap” multiple is partly an artifact of peak earnings that may already be gone. Revenue is contracting, not merely decelerating, and operating income has stopped growing even on a full-year basis. If the 500-plus basis points of margin loss is the new competitive equilibrium rather than a temporary integration or mix issue, then 8x trailing becomes 11–12x depressed earnings with negative growth—no longer a deep-value gift. The valuation synthesis’s $100 target leans on historical margins and mid-single-digit growth that the 2026 prints explicitly contradict; a smart opponent would say the models are rear-view and the 47% “discount” is rational repricing of a structurally impaired moat. I weigh that evidence seriously—it is why conviction is only moderate—but I do not let two soft quarters fully overwrite a multi-year FCF engine still throwing off $4B-plus with negligible leverage. The ice cube is melting slower than the price implies.
What would flip me: three more quarters of revenue declines and net margins stuck at or below 12% would force a re-rating to fairly valued or worse and collapse the FCF-yield argument; conversely, a return to positive mid-single-digit revenue growth and net margins re-expanding toward 15%+ in the next two reports would confirm the compression was transitory and justify a move toward the mid-80s or higher.
Big-3 Panel — where each AI stands
Advanced Analysis Forensic deep-dive · separate lenses
Fiserv shows the textbook profile of a mature, high-quality earner: revenue grew from 16.2B in 2021 to 21.2B in 2025 (roughly 7% CAGR), operating margin nearly doubled from 14.1% to 27.5%, and net income climbed from 1.33B to 3.48B. FCF has run 2.9B to 5.1B annually with OCF/NI of 2.08x and negative accruals (-3.2% of assets) - hallmarks of clean, cash-backed earnings. Diluted shares fell from 671.6M to 549.0M, a -4.9% CAGR, so per-share value is being concentrated meaningfully while SBC is modest at 1.7% of revenue. The blemish is the balance sheet: net debt of ~441M against only 798M of liquid cash (2.9% of market cap), short-term debt of 1.24B exceeding liquid cash, and Altman Z of 1.3 flagging distress on the model - though Altman is unreliable for asset-light processors carrying goodwill from the First Data merger. The lone insider buy (Fritz, ~520K open-market) is a small positive tell but not material. Overall this reads as a durable, well-run franchise where cash generation comfortably services leverage, but the debt load prevents a fortress rating.
Verify before trusting this (6)
- Total debt schedule and maturity ladder - what refinances in the next 12-24 months and at what rate
- Composition of the 1.24B short-term debt (commercial paper vs term maturities) and available revolver capacity
- Whether the 2024-to-2025 FCF step-down (5.06B to 4.30B) is working capital timing or a durable capex/tax shift
- Merchant/Clover segment growth vs legacy processing - is the margin expansion durable or mix-driven
- Customer concentration in Financial Solutions segment and any large contract renewals
- Goodwill/intangibles balance from First Data and any impairment testing sensitivity
The price anchor is $52.66 against a composite fair value of $95.87 and a signal-adjusted FV of $100.23, implying roughly 80-90% upside on the model's face. Both underlying methods corroborate: DCF at $101.82 and, more importantly, an EPV floor of $83.97 that does not require growth heroics. Even if I haircut the DCF for terminal-value optimism and lean on EPV, the deserved value sits near $80-85, still ~55-60% above spot. Earnings quality is flagged good (no haircut needed) and the Company-Quality lens grades the business Strong, which supports rather than inflates deserved value. The bear case (structural fintech disintermediation, margin compression) is coherent but would need to compress EPV meaningfully to justify $53; today's tape prices in a substantial slice of that decline already. The main honest caveat is leverage - net debt with short maturities means the equity FV is sensitive to rate/refi assumptions, and I would trim the DCF for that. Even so, price is well below a conservatively-derived deserved value, so the margin of safety is real, not manufactured.
Verify before trusting this (5)
- Organic revenue growth ex-Clover and merchant acceptance trend in latest 10-Q
- Segment margin trajectory - is the 14%-to-28% operating margin expansion still intact or plateauing
- Debt maturity ladder and refi rates - how much interest expense steps up in next 24 months
- Buyback pace vs SBC dilution - net share count reduction
- Any guidance cuts or client-loss disclosures from large FI customers
The macro tape is roughly neutral (VIX 16, S&P just 2% off highs), so market-wide pressure on this name is muted. But FISV-specific sentiment is clearly negative: the archetype is 'fallen-angel' with minimal intensity and fragile durability, meaning there is no active bull story defending the stock and the bear thesis (fintech disruption, margin compression, legacy plumbing) is the working narrative. The August 6 Q2 miss and 2026 guidance cut re-anchored expectations lower, and follow-on coverage (Zacks-style 'deep discount but weak growth', 2026 outlook cut pieces) keeps reinforcing the doubt.
Verify before trusting this (4)
- Whether sell-side estimate revisions stabilize or keep drifting lower into next print
- Any credible re-framing (AI-payments, agentic commerce role) that gives FISV a story rather than being a passive APA member
- Follow-through insider buying vs further activist trimming
- Sector rotation into defensive/quality-value that could lift forgotten compounders
AI reaches Fiserv through three channels of very different sign: on cost, a headcount-heavy service, implementation and support organization sitting behind volume-based pricing is a clean margin lever, and mid-20s to high-20s operating margins with $4-5B FCF show the structure can absorb it; on the revenue unit, transactions and accounts survive cheap intelligence almost entirely, which is why this is not a seat-count story; on defensibility, the exposure is real — AI compresses both the cost of building modern issuer/core software and the cost of migrating off Fiserv's legacy platforms, and at the merchant end agentic checkout could shift the value-added layer above Clover to wallet and agent platforms while leaving Fiserv the lower-margin processing. The observed 3.9% share gap versus industry growth predates AI and is mostly competitive execution, but AI is the mechanism that could make that gap structural rather than cyclical.
Verify before trusting this (8)
- Regulatory consent orders or exam findings
- Fraud loss absorption terms
- Compliance headcount vs revenue
- Yield per transaction
- Volume vs revenue growth spread
- Repricing at large renewals
- Financial Solutions organic growth
- Finxact conversion cadence
Payments infrastructure remains a growing, high-margin category, but the value is migrating from legacy processing plumbing to software-led, vertically-integrated acquirers and API-native platforms. Open banking and in-house/cloud core alternatives slowly erode the FI-side moat's pricing power rather than its retention. Macro adds a cyclical drag: high rates and softening SMB formation cap transaction volume growth just as inflation-linked and float-type revenue tailwinds reverse. Net: the world is not abandoning the function Fiserv performs, but it is repricing who gets paid for it.
When we made this prediction on Aug 21, 2026, FISV was $52.76. We expect it to be $66.50 by Feb 2027, and we consider it great value under $65.00. This is an early model (v0.6.0) — the direction is more reliable than the exact price. Made Aug 21, 2026.
Blue is our prediction, starting the day we made it. Grey is a slower route to the same place — the same destination, taking longer. Black is the actual price, so you can see how we are doing.